diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,25527 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.6415689021187183,
+  "eval_steps": 100,
+  "global_step": 37500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2000000000000002e-07,
+      "loss": 3.306,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 3.5052,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 3.3971,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6e-06,
+      "loss": 3.3713,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 3.4039,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6e-06,
+      "loss": 3.3735,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.2e-06,
+      "loss": 3.3117,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 3.2573,
+      "step": 70
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 3.1882,
+      "step": 80
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.08e-05,
+      "loss": 3.0751,
+      "step": 90
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2e-05,
+      "loss": 2.9993,
+      "step": 100
+    },
+    {
+      "epoch": 0.0,
+      "eval_loss": 2.979447841644287,
+      "eval_runtime": 11.7537,
+      "eval_samples_per_second": 348.487,
+      "eval_steps_per_second": 21.78,
+      "step": 100
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.32e-05,
+      "loss": 2.9849,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.44e-05,
+      "loss": 2.9377,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.56e-05,
+      "loss": 2.916,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6800000000000002e-05,
+      "loss": 2.8504,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8e-05,
+      "loss": 2.8581,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9200000000000003e-05,
+      "loss": 2.8143,
+      "step": 160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.04e-05,
+      "loss": 2.8103,
+      "step": 170
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.16e-05,
+      "loss": 2.824,
+      "step": 180
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2800000000000002e-05,
+      "loss": 2.7853,
+      "step": 190
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4e-05,
+      "loss": 2.7613,
+      "step": 200
+    },
+    {
+      "epoch": 0.01,
+      "eval_loss": 2.7339868545532227,
+      "eval_runtime": 11.8471,
+      "eval_samples_per_second": 345.74,
+      "eval_steps_per_second": 21.609,
+      "step": 200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.52e-05,
+      "loss": 2.7264,
+      "step": 210
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.64e-05,
+      "loss": 2.7563,
+      "step": 220
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7600000000000003e-05,
+      "loss": 2.7032,
+      "step": 230
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.88e-05,
+      "loss": 2.6709,
+      "step": 240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3e-05,
+      "loss": 2.7113,
+      "step": 250
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.12e-05,
+      "loss": 2.652,
+      "step": 260
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.24e-05,
+      "loss": 2.6599,
+      "step": 270
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3600000000000004e-05,
+      "loss": 2.6164,
+      "step": 280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.48e-05,
+      "loss": 2.6489,
+      "step": 290
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6e-05,
+      "loss": 2.6529,
+      "step": 300
+    },
+    {
+      "epoch": 0.01,
+      "eval_loss": 2.6033902168273926,
+      "eval_runtime": 11.9371,
+      "eval_samples_per_second": 343.131,
+      "eval_steps_per_second": 21.446,
+      "step": 300
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.72e-05,
+      "loss": 2.6186,
+      "step": 310
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8400000000000005e-05,
+      "loss": 2.6158,
+      "step": 320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.96e-05,
+      "loss": 2.5911,
+      "step": 330
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.08e-05,
+      "loss": 2.5856,
+      "step": 340
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2e-05,
+      "loss": 2.5734,
+      "step": 350
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.32e-05,
+      "loss": 2.6221,
+      "step": 360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.44e-05,
+      "loss": 2.5921,
+      "step": 370
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5600000000000004e-05,
+      "loss": 2.5796,
+      "step": 380
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6800000000000006e-05,
+      "loss": 2.5515,
+      "step": 390
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8e-05,
+      "loss": 2.5531,
+      "step": 400
+    },
+    {
+      "epoch": 0.02,
+      "eval_loss": 2.5212419033050537,
+      "eval_runtime": 11.6859,
+      "eval_samples_per_second": 350.507,
+      "eval_steps_per_second": 21.907,
+      "step": 400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9199999999999997e-05,
+      "loss": 2.5611,
+      "step": 410
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.04e-05,
+      "loss": 2.5204,
+      "step": 420
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.16e-05,
+      "loss": 2.5302,
+      "step": 430
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.28e-05,
+      "loss": 2.453,
+      "step": 440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.4000000000000005e-05,
+      "loss": 2.5275,
+      "step": 450
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.520000000000001e-05,
+      "loss": 2.4754,
+      "step": 460
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.6399999999999995e-05,
+      "loss": 2.5214,
+      "step": 470
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.76e-05,
+      "loss": 2.475,
+      "step": 480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.88e-05,
+      "loss": 2.4533,
+      "step": 490
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6e-05,
+      "loss": 2.4883,
+      "step": 500
+    },
+    {
+      "epoch": 0.02,
+      "eval_loss": 2.4650485515594482,
+      "eval_runtime": 11.6784,
+      "eval_samples_per_second": 350.734,
+      "eval_steps_per_second": 21.921,
+      "step": 500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.999999274988748e-05,
+      "loss": 2.4537,
+      "step": 510
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.9999970999553426e-05,
+      "loss": 2.488,
+      "step": 520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.9999934749008346e-05,
+      "loss": 2.5225,
+      "step": 530
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.999988399826977e-05,
+      "loss": 2.4376,
+      "step": 540
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.999981874736221e-05,
+      "loss": 2.4825,
+      "step": 550
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.9999738996317224e-05,
+      "loss": 2.4779,
+      "step": 560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.999964474517335e-05,
+      "loss": 2.4146,
+      "step": 570
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.9999535993976145e-05,
+      "loss": 2.4468,
+      "step": 580
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999941274277817e-05,
+      "loss": 2.4516,
+      "step": 590
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.9999274991639004e-05,
+      "loss": 2.4213,
+      "step": 600
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 2.411180019378662,
+      "eval_runtime": 11.7741,
+      "eval_samples_per_second": 347.881,
+      "eval_steps_per_second": 21.743,
+      "step": 600
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999912274062522e-05,
+      "loss": 2.4222,
+      "step": 610
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999895598981041e-05,
+      "loss": 2.3641,
+      "step": 620
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999877473927517e-05,
+      "loss": 2.4214,
+      "step": 630
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999857898910712e-05,
+      "loss": 2.3662,
+      "step": 640
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999836873940085e-05,
+      "loss": 2.4183,
+      "step": 650
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.9998143990258e-05,
+      "loss": 2.4012,
+      "step": 660
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.9997904741787194e-05,
+      "loss": 2.4289,
+      "step": 670
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999765099410407e-05,
+      "loss": 2.3593,
+      "step": 680
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999738274733128e-05,
+      "loss": 2.4153,
+      "step": 690
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999710000159846e-05,
+      "loss": 2.4046,
+      "step": 700
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 2.3694727420806885,
+      "eval_runtime": 12.1658,
+      "eval_samples_per_second": 336.681,
+      "eval_steps_per_second": 21.043,
+      "step": 700
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999680275704231e-05,
+      "loss": 2.3952,
+      "step": 710
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999649101380646e-05,
+      "loss": 2.381,
+      "step": 720
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999616477204161e-05,
+      "loss": 2.3835,
+      "step": 730
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.9995824031905446e-05,
+      "loss": 2.3664,
+      "step": 740
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.9995468793562655e-05,
+      "loss": 2.3001,
+      "step": 750
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.9995099057184944e-05,
+      "loss": 2.3884,
+      "step": 760
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999471482295101e-05,
+      "loss": 2.3915,
+      "step": 770
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999431609104658e-05,
+      "loss": 2.3674,
+      "step": 780
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999390286166438e-05,
+      "loss": 2.3697,
+      "step": 790
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.9993475135004135e-05,
+      "loss": 2.3509,
+      "step": 800
+    },
+    {
+      "epoch": 0.04,
+      "eval_loss": 2.3388166427612305,
+      "eval_runtime": 11.6279,
+      "eval_samples_per_second": 352.257,
+      "eval_steps_per_second": 22.016,
+      "step": 800
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.9993032911272584e-05,
+      "loss": 2.3633,
+      "step": 810
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.999257619068346e-05,
+      "loss": 2.3405,
+      "step": 820
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.9992104973457536e-05,
+      "loss": 2.3684,
+      "step": 830
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.9991619259822554e-05,
+      "loss": 2.3348,
+      "step": 840
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.999111905001329e-05,
+      "loss": 2.396,
+      "step": 850
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.999060434427151e-05,
+      "loss": 2.3006,
+      "step": 860
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.999007514284599e-05,
+      "loss": 2.3543,
+      "step": 870
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.998953144599253e-05,
+      "loss": 2.3564,
+      "step": 880
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.9988973253973895e-05,
+      "loss": 2.3265,
+      "step": 890
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.9988400567059905e-05,
+      "loss": 2.3153,
+      "step": 900
+    },
+    {
+      "epoch": 0.04,
+      "eval_loss": 2.309986114501953,
+      "eval_runtime": 11.6011,
+      "eval_samples_per_second": 353.071,
+      "eval_steps_per_second": 22.067,
+      "step": 900
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.9987813385527355e-05,
+      "loss": 2.2815,
+      "step": 910
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.998721170966004e-05,
+      "loss": 2.3601,
+      "step": 920
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.99865955397488e-05,
+      "loss": 2.3063,
+      "step": 930
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.998596487609144e-05,
+      "loss": 2.3125,
+      "step": 940
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.998531971899279e-05,
+      "loss": 2.2972,
+      "step": 950
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.9984660068764675e-05,
+      "loss": 2.2984,
+      "step": 960
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.998398592572594e-05,
+      "loss": 2.2664,
+      "step": 970
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.998329729020241e-05,
+      "loss": 2.3188,
+      "step": 980
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.998259416252695e-05,
+      "loss": 2.2963,
+      "step": 990
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.99818765430394e-05,
+      "loss": 2.3158,
+      "step": 1000
+    },
+    {
+      "epoch": 0.04,
+      "eval_loss": 2.2877612113952637,
+      "eval_runtime": 12.462,
+      "eval_samples_per_second": 328.678,
+      "eval_steps_per_second": 20.542,
+      "step": 1000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.998114443208661e-05,
+      "loss": 2.313,
+      "step": 1010
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.998039783002245e-05,
+      "loss": 2.3052,
+      "step": 1020
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.997963673720778e-05,
+      "loss": 2.3348,
+      "step": 1030
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.997886115401047e-05,
+      "loss": 2.3304,
+      "step": 1040
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.997807108080538e-05,
+      "loss": 2.2719,
+      "step": 1050
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.9977266517974396e-05,
+      "loss": 2.2937,
+      "step": 1060
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.9976447465906386e-05,
+      "loss": 2.2919,
+      "step": 1070
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.997561392499724e-05,
+      "loss": 2.2525,
+      "step": 1080
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.997476589564984e-05,
+      "loss": 2.2822,
+      "step": 1090
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.997390337827408e-05,
+      "loss": 2.2833,
+      "step": 1100
+    },
+    {
+      "epoch": 0.05,
+      "eval_loss": 2.267622947692871,
+      "eval_runtime": 11.9236,
+      "eval_samples_per_second": 343.521,
+      "eval_steps_per_second": 21.47,
+      "step": 1100
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.997302637328683e-05,
+      "loss": 2.2687,
+      "step": 1110
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.9972134881112e-05,
+      "loss": 2.2636,
+      "step": 1120
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.997122890218047e-05,
+      "loss": 2.3226,
+      "step": 1130
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.997030843693016e-05,
+      "loss": 2.2599,
+      "step": 1140
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.996937348580595e-05,
+      "loss": 2.2566,
+      "step": 1150
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.9968424049259735e-05,
+      "loss": 2.2997,
+      "step": 1160
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.9967460127750446e-05,
+      "loss": 2.3179,
+      "step": 1170
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.996648172174395e-05,
+      "loss": 2.2863,
+      "step": 1180
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.996548883171317e-05,
+      "loss": 2.267,
+      "step": 1190
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.9964481458138e-05,
+      "loss": 2.2207,
+      "step": 1200
+    },
+    {
+      "epoch": 0.05,
+      "eval_loss": 2.2525129318237305,
+      "eval_runtime": 12.0941,
+      "eval_samples_per_second": 338.679,
+      "eval_steps_per_second": 21.167,
+      "step": 1200
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.996345960150536e-05,
+      "loss": 2.2595,
+      "step": 1210
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.9962423262309146e-05,
+      "loss": 2.316,
+      "step": 1220
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.996137244105027e-05,
+      "loss": 2.258,
+      "step": 1230
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.996030713823661e-05,
+      "loss": 2.2726,
+      "step": 1240
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.995922735438311e-05,
+      "loss": 2.249,
+      "step": 1250
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.995813309001164e-05,
+      "loss": 2.2319,
+      "step": 1260
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.995702434565112e-05,
+      "loss": 2.2444,
+      "step": 1270
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.995590112183745e-05,
+      "loss": 2.258,
+      "step": 1280
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.995476341911353e-05,
+      "loss": 2.296,
+      "step": 1290
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.995361123802926e-05,
+      "loss": 2.2656,
+      "step": 1300
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 2.233938694000244,
+      "eval_runtime": 11.9568,
+      "eval_samples_per_second": 342.566,
+      "eval_steps_per_second": 21.41,
+      "step": 1300
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.995244457914152e-05,
+      "loss": 2.2022,
+      "step": 1310
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.9951263443014225e-05,
+      "loss": 2.2465,
+      "step": 1320
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.9950067830218256e-05,
+      "loss": 2.2018,
+      "step": 1330
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.99488577413315e-05,
+      "loss": 2.2699,
+      "step": 1340
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.994763317693885e-05,
+      "loss": 2.2657,
+      "step": 1350
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.994639413763218e-05,
+      "loss": 2.2484,
+      "step": 1360
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.994514062401038e-05,
+      "loss": 2.2652,
+      "step": 1370
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.9943872636679305e-05,
+      "loss": 2.2168,
+      "step": 1380
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.9942590176251846e-05,
+      "loss": 2.2863,
+      "step": 1390
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.9941293243347845e-05,
+      "loss": 2.1976,
+      "step": 1400
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 2.2203006744384766,
+      "eval_runtime": 16.8316,
+      "eval_samples_per_second": 243.352,
+      "eval_steps_per_second": 15.209,
+      "step": 1400
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.993998183859419e-05,
+      "loss": 2.2344,
+      "step": 1410
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.993865596262472e-05,
+      "loss": 2.2125,
+      "step": 1420
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.993731561608029e-05,
+      "loss": 2.2113,
+      "step": 1430
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.993596079960873e-05,
+      "loss": 2.2446,
+      "step": 1440
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.99345915138649e-05,
+      "loss": 2.2338,
+      "step": 1450
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.993320775951061e-05,
+      "loss": 2.247,
+      "step": 1460
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.9931809537214715e-05,
+      "loss": 2.1983,
+      "step": 1470
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.9930396847653e-05,
+      "loss": 2.2772,
+      "step": 1480
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.9928969691508296e-05,
+      "loss": 2.232,
+      "step": 1490
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.99275280694704e-05,
+      "loss": 2.2015,
+      "step": 1500
+    },
+    {
+      "epoch": 0.07,
+      "eval_loss": 2.207272529602051,
+      "eval_runtime": 11.8836,
+      "eval_samples_per_second": 344.677,
+      "eval_steps_per_second": 21.542,
+      "step": 1500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.9926071982236105e-05,
+      "loss": 2.2393,
+      "step": 1510
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.9924601430509207e-05,
+      "loss": 2.2068,
+      "step": 1520
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.992311641500047e-05,
+      "loss": 2.2571,
+      "step": 1530
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.992161693642767e-05,
+      "loss": 2.1754,
+      "step": 1540
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.9920102995515554e-05,
+      "loss": 2.2719,
+      "step": 1550
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.99185745929959e-05,
+      "loss": 2.1849,
+      "step": 1560
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.991703172960742e-05,
+      "loss": 2.2507,
+      "step": 1570
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.991547440609585e-05,
+      "loss": 2.238,
+      "step": 1580
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.991390262321392e-05,
+      "loss": 2.204,
+      "step": 1590
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.9912316381721316e-05,
+      "loss": 2.2177,
+      "step": 1600
+    },
+    {
+      "epoch": 0.07,
+      "eval_loss": 2.1963398456573486,
+      "eval_runtime": 13.5095,
+      "eval_samples_per_second": 303.193,
+      "eval_steps_per_second": 18.95,
+      "step": 1600
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.9910715682384755e-05,
+      "loss": 2.1927,
+      "step": 1610
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.990910052597791e-05,
+      "loss": 2.2564,
+      "step": 1620
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.990747091328144e-05,
+      "loss": 2.2149,
+      "step": 1630
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.990582684508302e-05,
+      "loss": 2.2022,
+      "step": 1640
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.99041683221773e-05,
+      "loss": 2.2009,
+      "step": 1650
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.99024953453659e-05,
+      "loss": 2.2003,
+      "step": 1660
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.9900807915457434e-05,
+      "loss": 2.2312,
+      "step": 1670
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.989910603326752e-05,
+      "loss": 2.1916,
+      "step": 1680
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.9897389699618725e-05,
+      "loss": 2.2396,
+      "step": 1690
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.9895658915340654e-05,
+      "loss": 2.175,
+      "step": 1700
+    },
+    {
+      "epoch": 0.07,
+      "eval_loss": 2.1844124794006348,
+      "eval_runtime": 11.4247,
+      "eval_samples_per_second": 358.521,
+      "eval_steps_per_second": 22.408,
+      "step": 1700
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.989391368126984e-05,
+      "loss": 2.1792,
+      "step": 1710
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.989215399824984e-05,
+      "loss": 2.2737,
+      "step": 1720
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.9890379867131177e-05,
+      "loss": 2.1745,
+      "step": 1730
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.9888591288771357e-05,
+      "loss": 2.1888,
+      "step": 1740
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.988678826403488e-05,
+      "loss": 2.1556,
+      "step": 1750
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.988497079379321e-05,
+      "loss": 2.2134,
+      "step": 1760
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.988313887892482e-05,
+      "loss": 2.1801,
+      "step": 1770
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.988129252031514e-05,
+      "loss": 2.1928,
+      "step": 1780
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.9879431718856594e-05,
+      "loss": 2.1882,
+      "step": 1790
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.987755647544857e-05,
+      "loss": 2.2182,
+      "step": 1800
+    },
+    {
+      "epoch": 0.08,
+      "eval_loss": 2.174042224884033,
+      "eval_runtime": 11.6019,
+      "eval_samples_per_second": 353.045,
+      "eval_steps_per_second": 22.065,
+      "step": 1800
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.987566679099746e-05,
+      "loss": 2.178,
+      "step": 1810
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.987376266641664e-05,
+      "loss": 2.2011,
+      "step": 1820
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.987184410262643e-05,
+      "loss": 2.2196,
+      "step": 1830
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.9869911100554165e-05,
+      "loss": 2.2287,
+      "step": 1840
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.986796366113413e-05,
+      "loss": 2.1855,
+      "step": 1850
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.986600178530761e-05,
+      "loss": 2.2516,
+      "step": 1860
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.9864025474022856e-05,
+      "loss": 2.1804,
+      "step": 1870
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.98620347282351e-05,
+      "loss": 2.2544,
+      "step": 1880
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.986002954890657e-05,
+      "loss": 2.1874,
+      "step": 1890
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.985800993700642e-05,
+      "loss": 2.1908,
+      "step": 1900
+    },
+    {
+      "epoch": 0.08,
+      "eval_loss": 2.1624016761779785,
+      "eval_runtime": 11.7335,
+      "eval_samples_per_second": 349.086,
+      "eval_steps_per_second": 21.818,
+      "step": 1900
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.985597589351083e-05,
+      "loss": 2.1858,
+      "step": 1910
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.985392741940292e-05,
+      "loss": 2.1944,
+      "step": 1920
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.985186451567282e-05,
+      "loss": 2.2135,
+      "step": 1930
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.984978718331762e-05,
+      "loss": 2.1289,
+      "step": 1940
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.9847695423341354e-05,
+      "loss": 2.1891,
+      "step": 1950
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.9845589236755065e-05,
+      "loss": 2.1651,
+      "step": 1960
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.984346862457677e-05,
+      "loss": 2.2016,
+      "step": 1970
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.984133358783144e-05,
+      "loss": 2.2092,
+      "step": 1980
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.983918412755103e-05,
+      "loss": 2.1864,
+      "step": 1990
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.983702024477445e-05,
+      "loss": 2.1916,
+      "step": 2000
+    },
+    {
+      "epoch": 0.09,
+      "eval_loss": 2.1545424461364746,
+      "eval_runtime": 12.2776,
+      "eval_samples_per_second": 333.616,
+      "eval_steps_per_second": 20.851,
+      "step": 2000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.9834841940547604e-05,
+      "loss": 2.1598,
+      "step": 2010
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.983264921592335e-05,
+      "loss": 2.1968,
+      "step": 2020
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.9830442071961524e-05,
+      "loss": 2.1737,
+      "step": 2030
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.982822050972892e-05,
+      "loss": 2.1664,
+      "step": 2040
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.982598453029933e-05,
+      "loss": 2.2088,
+      "step": 2050
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.982373413475348e-05,
+      "loss": 2.1314,
+      "step": 2060
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.9821469324179084e-05,
+      "loss": 2.1583,
+      "step": 2070
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.98191900996708e-05,
+      "loss": 2.2043,
+      "step": 2080
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.981689646233029e-05,
+      "loss": 2.1916,
+      "step": 2090
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.981458841326616e-05,
+      "loss": 2.0975,
+      "step": 2100
+    },
+    {
+      "epoch": 0.09,
+      "eval_loss": 2.145850896835327,
+      "eval_runtime": 11.7678,
+      "eval_samples_per_second": 348.07,
+      "eval_steps_per_second": 21.754,
+      "step": 2100
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.981226595359397e-05,
+      "loss": 2.1805,
+      "step": 2110
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.980992908443628e-05,
+      "loss": 2.1319,
+      "step": 2120
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.9807577806922587e-05,
+      "loss": 2.1684,
+      "step": 2130
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.980521212218935e-05,
+      "loss": 2.1669,
+      "step": 2140
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.9802832031380006e-05,
+      "loss": 2.1536,
+      "step": 2150
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.980043753564496e-05,
+      "loss": 2.1812,
+      "step": 2160
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.979802863614155e-05,
+      "loss": 2.1474,
+      "step": 2170
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.979560533403412e-05,
+      "loss": 2.1461,
+      "step": 2180
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.979316763049393e-05,
+      "loss": 2.1817,
+      "step": 2190
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.979071552669923e-05,
+      "loss": 2.2076,
+      "step": 2200
+    },
+    {
+      "epoch": 0.1,
+      "eval_loss": 2.136880397796631,
+      "eval_runtime": 11.5412,
+      "eval_samples_per_second": 354.904,
+      "eval_steps_per_second": 22.181,
+      "step": 2200
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.978824902383523e-05,
+      "loss": 2.1546,
+      "step": 2210
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.978576812309407e-05,
+      "loss": 2.1107,
+      "step": 2220
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.97832728256749e-05,
+      "loss": 2.1283,
+      "step": 2230
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.9780763132783765e-05,
+      "loss": 2.1638,
+      "step": 2240
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.977823904563373e-05,
+      "loss": 2.2048,
+      "step": 2250
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.9775700565444777e-05,
+      "loss": 2.1297,
+      "step": 2260
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.977314769344385e-05,
+      "loss": 2.1407,
+      "step": 2270
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.9770580430864876e-05,
+      "loss": 2.175,
+      "step": 2280
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.976799877894871e-05,
+      "loss": 2.1167,
+      "step": 2290
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.976540273894317e-05,
+      "loss": 2.1116,
+      "step": 2300
+    },
+    {
+      "epoch": 0.1,
+      "eval_loss": 2.1296005249023438,
+      "eval_runtime": 11.6525,
+      "eval_samples_per_second": 351.512,
+      "eval_steps_per_second": 21.97,
+      "step": 2300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.976279231210302e-05,
+      "loss": 2.1281,
+      "step": 2310
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.9760167499689985e-05,
+      "loss": 2.1152,
+      "step": 2320
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.9757528302972764e-05,
+      "loss": 2.1657,
+      "step": 2330
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.975487472322696e-05,
+      "loss": 2.1417,
+      "step": 2340
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.975220676173518e-05,
+      "loss": 2.1868,
+      "step": 2350
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.9749524419786954e-05,
+      "loss": 2.1341,
+      "step": 2360
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.974682769867876e-05,
+      "loss": 2.1344,
+      "step": 2370
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.974411659971403e-05,
+      "loss": 2.111,
+      "step": 2380
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.974139112420316e-05,
+      "loss": 2.1725,
+      "step": 2390
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.973865127346348e-05,
+      "loss": 2.1404,
+      "step": 2400
+    },
+    {
+      "epoch": 0.11,
+      "eval_loss": 2.1210813522338867,
+      "eval_runtime": 12.1931,
+      "eval_samples_per_second": 335.927,
+      "eval_steps_per_second": 20.995,
+      "step": 2400
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.973589704881927e-05,
+      "loss": 2.1281,
+      "step": 2410
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.9733128451601764e-05,
+      "loss": 2.1381,
+      "step": 2420
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.9730345483149134e-05,
+      "loss": 2.1297,
+      "step": 2430
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.9727548144806494e-05,
+      "loss": 2.1399,
+      "step": 2440
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.9724736437925926e-05,
+      "loss": 2.1495,
+      "step": 2450
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.9721910363866444e-05,
+      "loss": 2.094,
+      "step": 2460
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.971906992399399e-05,
+      "loss": 2.1428,
+      "step": 2470
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.9716215119681477e-05,
+      "loss": 2.1544,
+      "step": 2480
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.971334595230875e-05,
+      "loss": 2.1708,
+      "step": 2490
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.971046242326258e-05,
+      "loss": 2.1338,
+      "step": 2500
+    },
+    {
+      "epoch": 0.11,
+      "eval_loss": 2.1137373447418213,
+      "eval_runtime": 11.9925,
+      "eval_samples_per_second": 341.548,
+      "eval_steps_per_second": 21.347,
+      "step": 2500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.9707564533936704e-05,
+      "loss": 2.0854,
+      "step": 2510
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.970465228573179e-05,
+      "loss": 2.1676,
+      "step": 2520
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.9701725680055445e-05,
+      "loss": 2.1919,
+      "step": 2530
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.9698784718322226e-05,
+      "loss": 2.116,
+      "step": 2540
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.9695829401953615e-05,
+      "loss": 2.1112,
+      "step": 2550
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.969285973237802e-05,
+      "loss": 2.1726,
+      "step": 2560
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.968987571103083e-05,
+      "loss": 2.1494,
+      "step": 2570
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.968687733935432e-05,
+      "loss": 2.0896,
+      "step": 2580
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.9683864618797746e-05,
+      "loss": 2.1028,
+      "step": 2590
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.968083755081727e-05,
+      "loss": 2.0907,
+      "step": 2600
+    },
+    {
+      "epoch": 0.11,
+      "eval_loss": 2.1094017028808594,
+      "eval_runtime": 11.9254,
+      "eval_samples_per_second": 343.467,
+      "eval_steps_per_second": 21.467,
+      "step": 2600
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.967779613687599e-05,
+      "loss": 2.1917,
+      "step": 2610
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.967474037844396e-05,
+      "loss": 2.0942,
+      "step": 2620
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.967167027699815e-05,
+      "loss": 2.1257,
+      "step": 2630
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.9668585834022466e-05,
+      "loss": 2.1354,
+      "step": 2640
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.966548705100773e-05,
+      "loss": 2.1495,
+      "step": 2650
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.966237392945172e-05,
+      "loss": 2.1276,
+      "step": 2660
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.965924647085914e-05,
+      "loss": 2.09,
+      "step": 2670
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.965610467674162e-05,
+      "loss": 2.1088,
+      "step": 2680
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.9652948548617715e-05,
+      "loss": 2.1951,
+      "step": 2690
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.9649778088012896e-05,
+      "loss": 2.1079,
+      "step": 2700
+    },
+    {
+      "epoch": 0.12,
+      "eval_loss": 2.104461669921875,
+      "eval_runtime": 12.0537,
+      "eval_samples_per_second": 339.813,
+      "eval_steps_per_second": 21.238,
+      "step": 2700
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.9646593296459594e-05,
+      "loss": 2.128,
+      "step": 2710
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.9643394175497145e-05,
+      "loss": 2.092,
+      "step": 2720
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.9640180726671806e-05,
+      "loss": 2.1446,
+      "step": 2730
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.9636952951536776e-05,
+      "loss": 2.1721,
+      "step": 2740
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.963371085165217e-05,
+      "loss": 2.1262,
+      "step": 2750
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.963045442858503e-05,
+      "loss": 2.1791,
+      "step": 2760
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.962718368390931e-05,
+      "loss": 2.0893,
+      "step": 2770
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.96238986192059e-05,
+      "loss": 2.1119,
+      "step": 2780
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.96205992360626e-05,
+      "loss": 2.1215,
+      "step": 2790
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.961728553607415e-05,
+      "loss": 2.1327,
+      "step": 2800
+    },
+    {
+      "epoch": 0.12,
+      "eval_loss": 2.0953927040100098,
+      "eval_runtime": 11.7403,
+      "eval_samples_per_second": 348.885,
+      "eval_steps_per_second": 21.805,
+      "step": 2800
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.9613957520842177e-05,
+      "loss": 2.0981,
+      "step": 2810
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.961061519197527e-05,
+      "loss": 2.1383,
+      "step": 2820
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.96072585510889e-05,
+      "loss": 2.1195,
+      "step": 2830
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.9603887599805455e-05,
+      "loss": 2.1172,
+      "step": 2840
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.960050233975428e-05,
+      "loss": 2.0543,
+      "step": 2850
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.959710277257159e-05,
+      "loss": 2.1454,
+      "step": 2860
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.959368889990055e-05,
+      "loss": 2.095,
+      "step": 2870
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.959026072339121e-05,
+      "loss": 2.1166,
+      "step": 2880
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.9586818244700554e-05,
+      "loss": 2.1447,
+      "step": 2890
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.9583361465492475e-05,
+      "loss": 2.1123,
+      "step": 2900
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 2.090653419494629,
+      "eval_runtime": 11.7523,
+      "eval_samples_per_second": 348.527,
+      "eval_steps_per_second": 21.783,
+      "step": 2900
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.957989038743777e-05,
+      "loss": 2.0771,
+      "step": 2910
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.9576405012214155e-05,
+      "loss": 2.1296,
+      "step": 2920
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.957290534150625e-05,
+      "loss": 2.1512,
+      "step": 2930
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.9569391377005604e-05,
+      "loss": 2.0663,
+      "step": 2940
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.9565863120410637e-05,
+      "loss": 2.1126,
+      "step": 2950
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.956232057342672e-05,
+      "loss": 2.1432,
+      "step": 2960
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.95587637377661e-05,
+      "loss": 2.1382,
+      "step": 2970
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.955519261514794e-05,
+      "loss": 2.0925,
+      "step": 2980
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.955160720729831e-05,
+      "loss": 2.1153,
+      "step": 2990
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.9548007515950196e-05,
+      "loss": 2.0952,
+      "step": 3000
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 2.085477828979492,
+      "eval_runtime": 12.1839,
+      "eval_samples_per_second": 336.181,
+      "eval_steps_per_second": 21.011,
+      "step": 3000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.954439354284346e-05,
+      "loss": 2.1191,
+      "step": 3010
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.954076528972489e-05,
+      "loss": 2.1233,
+      "step": 3020
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.953712275834817e-05,
+      "loss": 2.1523,
+      "step": 3030
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.953346595047388e-05,
+      "loss": 2.092,
+      "step": 3040
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.95297948678695e-05,
+      "loss": 2.0483,
+      "step": 3050
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.9526109512309423e-05,
+      "loss": 2.1245,
+      "step": 3060
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.9522409885574934e-05,
+      "loss": 2.1205,
+      "step": 3070
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.95186959894542e-05,
+      "loss": 2.1143,
+      "step": 3080
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.9514967825742325e-05,
+      "loss": 2.0988,
+      "step": 3090
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.9511225396241256e-05,
+      "loss": 2.1173,
+      "step": 3100
+    },
+    {
+      "epoch": 0.14,
+      "eval_loss": 2.0793402194976807,
+      "eval_runtime": 11.7138,
+      "eval_samples_per_second": 349.672,
+      "eval_steps_per_second": 21.854,
+      "step": 3100
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.950746870275986e-05,
+      "loss": 2.1051,
+      "step": 3110
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.950369774711392e-05,
+      "loss": 2.071,
+      "step": 3120
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.9499912531126096e-05,
+      "loss": 2.136,
+      "step": 3130
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.949611305662592e-05,
+      "loss": 2.1015,
+      "step": 3140
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.949229932544983e-05,
+      "loss": 2.1133,
+      "step": 3150
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.9488471339441175e-05,
+      "loss": 2.0993,
+      "step": 3160
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.948462910045017e-05,
+      "loss": 2.0734,
+      "step": 3170
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.948077261033392e-05,
+      "loss": 2.0588,
+      "step": 3180
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.9476901870956426e-05,
+      "loss": 2.0703,
+      "step": 3190
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.947301688418859e-05,
+      "loss": 2.0886,
+      "step": 3200
+    },
+    {
+      "epoch": 0.14,
+      "eval_loss": 2.0747480392456055,
+      "eval_runtime": 16.3253,
+      "eval_samples_per_second": 250.899,
+      "eval_steps_per_second": 15.681,
+      "step": 3200
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.9469117651908163e-05,
+      "loss": 2.0638,
+      "step": 3210
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.946520417599982e-05,
+      "loss": 2.0492,
+      "step": 3220
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.946127645835509e-05,
+      "loss": 2.0683,
+      "step": 3230
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.945733450087242e-05,
+      "loss": 2.0929,
+      "step": 3240
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.9453378305457085e-05,
+      "loss": 2.139,
+      "step": 3250
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.944940787402131e-05,
+      "loss": 2.0932,
+      "step": 3260
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.944542320848414e-05,
+      "loss": 2.1399,
+      "step": 3270
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.9441424310771545e-05,
+      "loss": 2.1349,
+      "step": 3280
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.943741118281635e-05,
+      "loss": 2.0773,
+      "step": 3290
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.943338382655826e-05,
+      "loss": 2.0418,
+      "step": 3300
+    },
+    {
+      "epoch": 0.14,
+      "eval_loss": 2.069995403289795,
+      "eval_runtime": 13.501,
+      "eval_samples_per_second": 303.384,
+      "eval_steps_per_second": 18.962,
+      "step": 3300
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.942934224394387e-05,
+      "loss": 2.1208,
+      "step": 3310
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.9425286436926635e-05,
+      "loss": 2.0428,
+      "step": 3320
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.942121640746688e-05,
+      "loss": 2.0581,
+      "step": 3330
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.941713215753184e-05,
+      "loss": 2.1182,
+      "step": 3340
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.9413033689095596e-05,
+      "loss": 2.098,
+      "step": 3350
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.940892100413909e-05,
+      "loss": 2.0887,
+      "step": 3360
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.9404794104650156e-05,
+      "loss": 2.0979,
+      "step": 3370
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.9400652992623495e-05,
+      "loss": 2.083,
+      "step": 3380
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.9396497670060685e-05,
+      "loss": 2.0856,
+      "step": 3390
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.939232813897014e-05,
+      "loss": 2.095,
+      "step": 3400
+    },
+    {
+      "epoch": 0.15,
+      "eval_loss": 2.0656051635742188,
+      "eval_runtime": 13.8953,
+      "eval_samples_per_second": 294.777,
+      "eval_steps_per_second": 18.424,
+      "step": 3400
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.9388144401367194e-05,
+      "loss": 2.1035,
+      "step": 3410
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.9383946459273995e-05,
+      "loss": 2.096,
+      "step": 3420
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.937973431471959e-05,
+      "loss": 2.1298,
+      "step": 3430
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.937550796973988e-05,
+      "loss": 2.0599,
+      "step": 3440
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.9371267426377624e-05,
+      "loss": 2.0537,
+      "step": 3450
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.9367012686682464e-05,
+      "loss": 2.0865,
+      "step": 3460
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.936274375271087e-05,
+      "loss": 2.1082,
+      "step": 3470
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.935846062652621e-05,
+      "loss": 2.1416,
+      "step": 3480
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.935416331019868e-05,
+      "loss": 2.0887,
+      "step": 3490
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.934985180580536e-05,
+      "loss": 2.0988,
+      "step": 3500
+    },
+    {
+      "epoch": 0.15,
+      "eval_loss": 2.061530113220215,
+      "eval_runtime": 14.5323,
+      "eval_samples_per_second": 281.854,
+      "eval_steps_per_second": 17.616,
+      "step": 3500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.934552611543016e-05,
+      "loss": 2.034,
+      "step": 3510
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.9341186241163875e-05,
+      "loss": 2.1009,
+      "step": 3520
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.9336832185104155e-05,
+      "loss": 2.1099,
+      "step": 3530
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.933246394935546e-05,
+      "loss": 2.0687,
+      "step": 3540
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.9328081536029175e-05,
+      "loss": 2.0504,
+      "step": 3550
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.9323684947243476e-05,
+      "loss": 2.0703,
+      "step": 3560
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.931927418512341e-05,
+      "loss": 2.0398,
+      "step": 3570
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.931484925180089e-05,
+      "loss": 2.1021,
+      "step": 3580
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.931041014941467e-05,
+      "loss": 2.0889,
+      "step": 3590
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.930595688011034e-05,
+      "loss": 2.092,
+      "step": 3600
+    },
+    {
+      "epoch": 0.16,
+      "eval_loss": 2.0562427043914795,
+      "eval_runtime": 11.8426,
+      "eval_samples_per_second": 345.869,
+      "eval_steps_per_second": 21.617,
+      "step": 3600
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.930148944604035e-05,
+      "loss": 2.0708,
+      "step": 3610
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.929700784936399e-05,
+      "loss": 2.0789,
+      "step": 3620
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.9292512092247405e-05,
+      "loss": 2.0735,
+      "step": 3630
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.928800217686358e-05,
+      "loss": 2.1018,
+      "step": 3640
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.928347810539233e-05,
+      "loss": 2.0637,
+      "step": 3650
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.927893988002033e-05,
+      "loss": 2.0389,
+      "step": 3660
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.9274387502941094e-05,
+      "loss": 2.062,
+      "step": 3670
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.926982097635497e-05,
+      "loss": 2.0766,
+      "step": 3680
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.926524030246914e-05,
+      "loss": 2.0654,
+      "step": 3690
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.926064548349764e-05,
+      "loss": 2.1046,
+      "step": 3700
+    },
+    {
+      "epoch": 0.16,
+      "eval_loss": 2.051886558532715,
+      "eval_runtime": 13.6021,
+      "eval_samples_per_second": 301.131,
+      "eval_steps_per_second": 18.821,
+      "step": 3700
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.9256036521661316e-05,
+      "loss": 2.068,
+      "step": 3710
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.9251413419187893e-05,
+      "loss": 2.0482,
+      "step": 3720
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.9246776178311885e-05,
+      "loss": 2.0652,
+      "step": 3730
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.9242124801274674e-05,
+      "loss": 2.1223,
+      "step": 3740
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.9237459290324444e-05,
+      "loss": 2.0723,
+      "step": 3750
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.9232779647716244e-05,
+      "loss": 2.054,
+      "step": 3760
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.9228085875711926e-05,
+      "loss": 2.1379,
+      "step": 3770
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.922337797658018e-05,
+      "loss": 2.0366,
+      "step": 3780
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.921865595259654e-05,
+      "loss": 2.0602,
+      "step": 3790
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.921391980604335e-05,
+      "loss": 2.0698,
+      "step": 3800
+    },
+    {
+      "epoch": 0.17,
+      "eval_loss": 2.047264814376831,
+      "eval_runtime": 15.2622,
+      "eval_samples_per_second": 268.375,
+      "eval_steps_per_second": 16.773,
+      "step": 3800
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.920916953920976e-05,
+      "loss": 2.0555,
+      "step": 3810
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.920440515439179e-05,
+      "loss": 2.121,
+      "step": 3820
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.9199626653892264e-05,
+      "loss": 2.0358,
+      "step": 3830
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.919483404002081e-05,
+      "loss": 2.0702,
+      "step": 3840
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.91900273150939e-05,
+      "loss": 2.0706,
+      "step": 3850
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.918520648143482e-05,
+      "loss": 2.0742,
+      "step": 3860
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.918037154137369e-05,
+      "loss": 2.0806,
+      "step": 3870
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.917552249724742e-05,
+      "loss": 2.0539,
+      "step": 3880
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.917065935139975e-05,
+      "loss": 2.0925,
+      "step": 3890
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.9165782106181244e-05,
+      "loss": 2.0628,
+      "step": 3900
+    },
+    {
+      "epoch": 0.17,
+      "eval_loss": 2.041888952255249,
+      "eval_runtime": 12.8465,
+      "eval_samples_per_second": 318.843,
+      "eval_steps_per_second": 19.928,
+      "step": 3900
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.916089076394927e-05,
+      "loss": 2.0655,
+      "step": 3910
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.915598532706801e-05,
+      "loss": 2.0998,
+      "step": 3920
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.915106579790848e-05,
+      "loss": 2.0661,
+      "step": 3930
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.914613217884846e-05,
+      "loss": 2.0268,
+      "step": 3940
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.9141184472272596e-05,
+      "loss": 2.1178,
+      "step": 3950
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.913622268057229e-05,
+      "loss": 2.0453,
+      "step": 3960
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.913124680614581e-05,
+      "loss": 2.0846,
+      "step": 3970
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.912625685139818e-05,
+      "loss": 2.1052,
+      "step": 3980
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.912125281874125e-05,
+      "loss": 2.0422,
+      "step": 3990
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.9116234710593674e-05,
+      "loss": 2.0528,
+      "step": 4000
+    },
+    {
+      "epoch": 0.18,
+      "eval_loss": 2.0398635864257812,
+      "eval_runtime": 12.2343,
+      "eval_samples_per_second": 334.796,
+      "eval_steps_per_second": 20.925,
+      "step": 4000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.9111202529380904e-05,
+      "loss": 2.0678,
+      "step": 4010
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.910615627753521e-05,
+      "loss": 2.0423,
+      "step": 4020
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.9101095957495654e-05,
+      "loss": 2.0402,
+      "step": 4030
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.909602157170808e-05,
+      "loss": 2.0706,
+      "step": 4040
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.909093312262516e-05,
+      "loss": 2.0197,
+      "step": 4050
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.9085830612706336e-05,
+      "loss": 2.0763,
+      "step": 4060
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.908071404441787e-05,
+      "loss": 2.1287,
+      "step": 4070
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.9075583420232806e-05,
+      "loss": 2.0297,
+      "step": 4080
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.907043874263098e-05,
+      "loss": 2.0862,
+      "step": 4090
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.906528001409902e-05,
+      "loss": 2.0211,
+      "step": 4100
+    },
+    {
+      "epoch": 0.18,
+      "eval_loss": 2.0350661277770996,
+      "eval_runtime": 12.0443,
+      "eval_samples_per_second": 340.079,
+      "eval_steps_per_second": 21.255,
+      "step": 4100
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.906010723713038e-05,
+      "loss": 2.0536,
+      "step": 4110
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.905492041422523e-05,
+      "loss": 2.0348,
+      "step": 4120
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.9049719547890594e-05,
+      "loss": 1.9906,
+      "step": 4130
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.9044504640640276e-05,
+      "loss": 2.0626,
+      "step": 4140
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.9039275694994844e-05,
+      "loss": 2.0877,
+      "step": 4150
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.9034032713481654e-05,
+      "loss": 2.0266,
+      "step": 4160
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.902877569863485e-05,
+      "loss": 2.13,
+      "step": 4170
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.9023504652995384e-05,
+      "loss": 2.0176,
+      "step": 4180
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.901821957911095e-05,
+      "loss": 2.0459,
+      "step": 4190
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.9012920479536034e-05,
+      "loss": 2.0805,
+      "step": 4200
+    },
+    {
+      "epoch": 0.18,
+      "eval_loss": 2.0328736305236816,
+      "eval_runtime": 18.7677,
+      "eval_samples_per_second": 218.248,
+      "eval_steps_per_second": 13.64,
+      "step": 4200
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.9007607356831934e-05,
+      "loss": 2.0437,
+      "step": 4210
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.900228021356666e-05,
+      "loss": 2.0958,
+      "step": 4220
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.899693905231507e-05,
+      "loss": 2.0722,
+      "step": 4230
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.899158387565877e-05,
+      "loss": 2.0361,
+      "step": 4240
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.89862146861861e-05,
+      "loss": 2.0838,
+      "step": 4250
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.898083148649224e-05,
+      "loss": 2.0631,
+      "step": 4260
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.8975434279179096e-05,
+      "loss": 2.0552,
+      "step": 4270
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.897002306685536e-05,
+      "loss": 2.0305,
+      "step": 4280
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.89645978521365e-05,
+      "loss": 2.1161,
+      "step": 4290
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.895915863764473e-05,
+      "loss": 2.0235,
+      "step": 4300
+    },
+    {
+      "epoch": 0.19,
+      "eval_loss": 2.0307788848876953,
+      "eval_runtime": 11.5851,
+      "eval_samples_per_second": 353.557,
+      "eval_steps_per_second": 22.097,
+      "step": 4300
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.895370542600906e-05,
+      "loss": 2.0678,
+      "step": 4310
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.894823821986524e-05,
+      "loss": 2.0671,
+      "step": 4320
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.894275702185579e-05,
+      "loss": 2.0524,
+      "step": 4330
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.893726183463001e-05,
+      "loss": 2.0369,
+      "step": 4340
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.893175266084394e-05,
+      "loss": 2.0658,
+      "step": 4350
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.892622950316039e-05,
+      "loss": 2.0488,
+      "step": 4360
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.8920692364248926e-05,
+      "loss": 2.0644,
+      "step": 4370
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.8915141246785875e-05,
+      "loss": 2.0313,
+      "step": 4380
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.890957615345433e-05,
+      "loss": 2.0117,
+      "step": 4390
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.89039970869441e-05,
+      "loss": 2.0123,
+      "step": 4400
+    },
+    {
+      "epoch": 0.19,
+      "eval_loss": 2.027006149291992,
+      "eval_runtime": 20.8382,
+      "eval_samples_per_second": 196.562,
+      "eval_steps_per_second": 12.285,
+      "step": 4400
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.88984040499518e-05,
+      "loss": 2.0617,
+      "step": 4410
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.889279704518077e-05,
+      "loss": 2.0394,
+      "step": 4420
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.888717607534109e-05,
+      "loss": 2.058,
+      "step": 4430
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.888154114314961e-05,
+      "loss": 2.1012,
+      "step": 4440
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.887589225132994e-05,
+      "loss": 2.0424,
+      "step": 4450
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.887022940261241e-05,
+      "loss": 2.0011,
+      "step": 4460
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.886455259973408e-05,
+      "loss": 2.1042,
+      "step": 4470
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.8858861845438824e-05,
+      "loss": 2.0368,
+      "step": 4480
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.885315714247719e-05,
+      "loss": 2.0797,
+      "step": 4490
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.8847438493606504e-05,
+      "loss": 2.0964,
+      "step": 4500
+    },
+    {
+      "epoch": 0.2,
+      "eval_loss": 2.023592710494995,
+      "eval_runtime": 13.8773,
+      "eval_samples_per_second": 295.157,
+      "eval_steps_per_second": 18.447,
+      "step": 4500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.884170590159081e-05,
+      "loss": 2.079,
+      "step": 4510
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.8835959369200925e-05,
+      "loss": 2.0535,
+      "step": 4520
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.883019889921436e-05,
+      "loss": 2.0473,
+      "step": 4530
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.882442449441539e-05,
+      "loss": 2.0421,
+      "step": 4540
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.881863615759503e-05,
+      "loss": 2.044,
+      "step": 4550
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.8812833891551024e-05,
+      "loss": 2.076,
+      "step": 4560
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.880701769908782e-05,
+      "loss": 2.0583,
+      "step": 4570
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.880118758301665e-05,
+      "loss": 2.0261,
+      "step": 4580
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.879534354615543e-05,
+      "loss": 2.0434,
+      "step": 4590
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.878948559132882e-05,
+      "loss": 2.0839,
+      "step": 4600
+    },
+    {
+      "epoch": 0.2,
+      "eval_loss": 2.0206034183502197,
+      "eval_runtime": 13.1899,
+      "eval_samples_per_second": 310.54,
+      "eval_steps_per_second": 19.409,
+      "step": 4600
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.8783613721368216e-05,
+      "loss": 2.0191,
+      "step": 4610
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.877772793911173e-05,
+      "loss": 2.037,
+      "step": 4620
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.8771828247404204e-05,
+      "loss": 2.0319,
+      "step": 4630
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.87659146490972e-05,
+      "loss": 2.0892,
+      "step": 4640
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.8759987147048997e-05,
+      "loss": 2.0116,
+      "step": 4650
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.87540457441246e-05,
+      "loss": 2.0725,
+      "step": 4660
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.874809044319573e-05,
+      "loss": 2.0583,
+      "step": 4670
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.8742121247140835e-05,
+      "loss": 2.0341,
+      "step": 4680
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.873613815884506e-05,
+      "loss": 2.0719,
+      "step": 4690
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.8730141181200284e-05,
+      "loss": 2.0736,
+      "step": 4700
+    },
+    {
+      "epoch": 0.21,
+      "eval_loss": 2.0166029930114746,
+      "eval_runtime": 13.4827,
+      "eval_samples_per_second": 303.798,
+      "eval_steps_per_second": 18.987,
+      "step": 4700
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.872413031710509e-05,
+      "loss": 2.0566,
+      "step": 4710
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.871810556946478e-05,
+      "loss": 2.0676,
+      "step": 4720
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.871206694119134e-05,
+      "loss": 2.0385,
+      "step": 4730
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.870601443520351e-05,
+      "loss": 2.0604,
+      "step": 4740
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.869994805442669e-05,
+      "loss": 2.0338,
+      "step": 4750
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.869386780179303e-05,
+      "loss": 1.9712,
+      "step": 4760
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.868777368024136e-05,
+      "loss": 2.0208,
+      "step": 4770
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.868166569271721e-05,
+      "loss": 2.0479,
+      "step": 4780
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.867554384217282e-05,
+      "loss": 2.0104,
+      "step": 4790
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.866940813156714e-05,
+      "loss": 2.0591,
+      "step": 4800
+    },
+    {
+      "epoch": 0.21,
+      "eval_loss": 2.0134992599487305,
+      "eval_runtime": 11.7962,
+      "eval_samples_per_second": 347.232,
+      "eval_steps_per_second": 21.702,
+      "step": 4800
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.8663258563865804e-05,
+      "loss": 2.041,
+      "step": 4810
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.865709514204115e-05,
+      "loss": 1.9956,
+      "step": 4820
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.865091786907221e-05,
+      "loss": 2.0447,
+      "step": 4830
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.864472674794471e-05,
+      "loss": 2.022,
+      "step": 4840
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.863852178165108e-05,
+      "loss": 1.9585,
+      "step": 4850
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.8632302973190424e-05,
+      "loss": 2.058,
+      "step": 4860
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.862607032556854e-05,
+      "loss": 2.0582,
+      "step": 4870
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.861982384179794e-05,
+      "loss": 2.0257,
+      "step": 4880
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.86135635248978e-05,
+      "loss": 1.9781,
+      "step": 4890
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.860728937789398e-05,
+      "loss": 2.0381,
+      "step": 4900
+    },
+    {
+      "epoch": 0.21,
+      "eval_loss": 2.0108816623687744,
+      "eval_runtime": 11.6358,
+      "eval_samples_per_second": 352.018,
+      "eval_steps_per_second": 22.001,
+      "step": 4900
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.860100140381903e-05,
+      "loss": 2.0079,
+      "step": 4910
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.8594699605712184e-05,
+      "loss": 2.0501,
+      "step": 4920
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.858838398661938e-05,
+      "loss": 2.0088,
+      "step": 4930
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.8582054549593184e-05,
+      "loss": 1.9654,
+      "step": 4940
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.8575711297692886e-05,
+      "loss": 2.0455,
+      "step": 4950
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.8569354233984445e-05,
+      "loss": 2.0596,
+      "step": 4960
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.856298336154048e-05,
+      "loss": 1.9902,
+      "step": 4970
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.855659868344029e-05,
+      "loss": 1.9897,
+      "step": 4980
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.8550200202769856e-05,
+      "loss": 2.0454,
+      "step": 4990
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.854378792262183e-05,
+      "loss": 2.0479,
+      "step": 5000
+    },
+    {
+      "epoch": 0.22,
+      "eval_loss": 2.0080699920654297,
+      "eval_runtime": 11.7197,
+      "eval_samples_per_second": 349.497,
+      "eval_steps_per_second": 21.844,
+      "step": 5000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.853736184609553e-05,
+      "loss": 2.037,
+      "step": 5010
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.853092197629693e-05,
+      "loss": 1.9935,
+      "step": 5020
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.852446831633869e-05,
+      "loss": 2.0225,
+      "step": 5030
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.851800086934013e-05,
+      "loss": 2.0627,
+      "step": 5040
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.851151963842721e-05,
+      "loss": 2.0333,
+      "step": 5050
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.85050246267326e-05,
+      "loss": 2.0602,
+      "step": 5060
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.849851583739559e-05,
+      "loss": 2.0367,
+      "step": 5070
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.849199327356215e-05,
+      "loss": 2.0453,
+      "step": 5080
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.84854569383849e-05,
+      "loss": 2.0275,
+      "step": 5090
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.847890683502312e-05,
+      "loss": 2.0145,
+      "step": 5100
+    },
+    {
+      "epoch": 0.22,
+      "eval_loss": 2.0048532485961914,
+      "eval_runtime": 11.7026,
+      "eval_samples_per_second": 350.008,
+      "eval_steps_per_second": 21.875,
+      "step": 5100
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.8472342966642735e-05,
+      "loss": 2.0088,
+      "step": 5110
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.8465765336416336e-05,
+      "loss": 2.0097,
+      "step": 5120
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.8459173947523165e-05,
+      "loss": 2.002,
+      "step": 5130
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.84525688031491e-05,
+      "loss": 2.007,
+      "step": 5140
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.8445949906486674e-05,
+      "loss": 2.0457,
+      "step": 5150
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.843931726073509e-05,
+      "loss": 2.0467,
+      "step": 5160
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.843267086910015e-05,
+      "loss": 1.9947,
+      "step": 5170
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.8426010734794346e-05,
+      "loss": 2.0507,
+      "step": 5180
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.84193368610368e-05,
+      "loss": 2.0486,
+      "step": 5190
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.841264925105323e-05,
+      "loss": 2.0231,
+      "step": 5200
+    },
+    {
+      "epoch": 0.23,
+      "eval_loss": 2.0028250217437744,
+      "eval_runtime": 11.7989,
+      "eval_samples_per_second": 347.151,
+      "eval_steps_per_second": 21.697,
+      "step": 5200
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.840594790807607e-05,
+      "loss": 1.9965,
+      "step": 5210
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.8399232835344335e-05,
+      "loss": 1.9954,
+      "step": 5220
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.83925040361037e-05,
+      "loss": 2.0133,
+      "step": 5230
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.838576151360646e-05,
+      "loss": 1.9867,
+      "step": 5240
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.837900527111156e-05,
+      "loss": 2.0275,
+      "step": 5250
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.837223531188456e-05,
+      "loss": 2.0276,
+      "step": 5260
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.836545163919767e-05,
+      "loss": 2.0017,
+      "step": 5270
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.83586542563297e-05,
+      "loss": 2.0644,
+      "step": 5280
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.835184316656612e-05,
+      "loss": 1.9864,
+      "step": 5290
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.834501837319899e-05,
+      "loss": 2.0212,
+      "step": 5300
+    },
+    {
+      "epoch": 0.23,
+      "eval_loss": 2.001617670059204,
+      "eval_runtime": 11.8117,
+      "eval_samples_per_second": 346.776,
+      "eval_steps_per_second": 21.673,
+      "step": 5300
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.8338179879527034e-05,
+      "loss": 2.0293,
+      "step": 5310
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.833132768885555e-05,
+      "loss": 2.0081,
+      "step": 5320
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.83244618044965e-05,
+      "loss": 2.0506,
+      "step": 5330
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.831758222976843e-05,
+      "loss": 2.0314,
+      "step": 5340
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.8310688967996534e-05,
+      "loss": 2.0002,
+      "step": 5350
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.830378202251261e-05,
+      "loss": 2.0547,
+      "step": 5360
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.829686139665505e-05,
+      "loss": 2.006,
+      "step": 5370
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.8289927093768885e-05,
+      "loss": 2.0277,
+      "step": 5380
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.8282979117205745e-05,
+      "loss": 2.0159,
+      "step": 5390
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.827601747032387e-05,
+      "loss": 2.0505,
+      "step": 5400
+    },
+    {
+      "epoch": 0.24,
+      "eval_loss": 1.9988398551940918,
+      "eval_runtime": 11.6914,
+      "eval_samples_per_second": 350.342,
+      "eval_steps_per_second": 21.896,
+      "step": 5400
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.8269042156488106e-05,
+      "loss": 2.0138,
+      "step": 5410
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.826205317906991e-05,
+      "loss": 2.0205,
+      "step": 5420
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.825505054144735e-05,
+      "loss": 2.0279,
+      "step": 5430
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.8248034247005075e-05,
+      "loss": 2.0387,
+      "step": 5440
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.8241004299134345e-05,
+      "loss": 2.0103,
+      "step": 5450
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.8233960701233026e-05,
+      "loss": 1.9852,
+      "step": 5460
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.822690345670558e-05,
+      "loss": 2.0353,
+      "step": 5470
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.821983256896305e-05,
+      "loss": 2.0069,
+      "step": 5480
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.821274804142309e-05,
+      "loss": 2.0343,
+      "step": 5490
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.820564987750994e-05,
+      "loss": 1.9844,
+      "step": 5500
+    },
+    {
+      "epoch": 0.24,
+      "eval_loss": 1.9970853328704834,
+      "eval_runtime": 11.77,
+      "eval_samples_per_second": 348.004,
+      "eval_steps_per_second": 21.75,
+      "step": 5500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.8198538080654456e-05,
+      "loss": 2.0359,
+      "step": 5510
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.819141265429402e-05,
+      "loss": 2.05,
+      "step": 5520
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.818427360187267e-05,
+      "loss": 1.985,
+      "step": 5530
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.817712092684099e-05,
+      "loss": 2.005,
+      "step": 5540
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.816995463265615e-05,
+      "loss": 2.0303,
+      "step": 5550
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.816277472278194e-05,
+      "loss": 2.0114,
+      "step": 5560
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.815558120068868e-05,
+      "loss": 2.0188,
+      "step": 5570
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.814837406985331e-05,
+      "loss": 2.0357,
+      "step": 5580
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.8141153333759316e-05,
+      "loss": 2.0272,
+      "step": 5590
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.8133918995896784e-05,
+      "loss": 2.0311,
+      "step": 5600
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 1.993647575378418,
+      "eval_runtime": 13.7808,
+      "eval_samples_per_second": 297.225,
+      "eval_steps_per_second": 18.577,
+      "step": 5600
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.8126671059762356e-05,
+      "loss": 2.0287,
+      "step": 5610
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.811940952885927e-05,
+      "loss": 2.0618,
+      "step": 5620
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.81121344066973e-05,
+      "loss": 2.025,
+      "step": 5630
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.810484569679283e-05,
+      "loss": 1.9983,
+      "step": 5640
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.8097543402668785e-05,
+      "loss": 1.9824,
+      "step": 5650
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.809022752785465e-05,
+      "loss": 1.9815,
+      "step": 5660
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.8082898075886495e-05,
+      "loss": 2.0088,
+      "step": 5670
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.807555505030695e-05,
+      "loss": 2.0073,
+      "step": 5680
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.80681984546652e-05,
+      "loss": 1.9424,
+      "step": 5690
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.806082829251696e-05,
+      "loss": 2.0114,
+      "step": 5700
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 1.9904594421386719,
+      "eval_runtime": 12.2556,
+      "eval_samples_per_second": 334.213,
+      "eval_steps_per_second": 20.888,
+      "step": 5700
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.805344456742457e-05,
+      "loss": 2.015,
+      "step": 5710
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.804604728295686e-05,
+      "loss": 2.0413,
+      "step": 5720
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.803863644268925e-05,
+      "loss": 2.0054,
+      "step": 5730
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.803121205020369e-05,
+      "loss": 1.9972,
+      "step": 5740
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.802377410908871e-05,
+      "loss": 2.007,
+      "step": 5750
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.801632262293935e-05,
+      "loss": 1.9748,
+      "step": 5760
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.8008857595357234e-05,
+      "loss": 2.0488,
+      "step": 5770
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.8001379029950515e-05,
+      "loss": 2.018,
+      "step": 5780
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.799388693033387e-05,
+      "loss": 1.9982,
+      "step": 5790
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.798638130012856e-05,
+      "loss": 2.0603,
+      "step": 5800
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 1.9879425764083862,
+      "eval_runtime": 19.1497,
+      "eval_samples_per_second": 213.894,
+      "eval_steps_per_second": 13.368,
+      "step": 5800
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.7978862142962354e-05,
+      "loss": 2.023,
+      "step": 5810
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.7971329462469565e-05,
+      "loss": 2.0262,
+      "step": 5820
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.7963783262291046e-05,
+      "loss": 1.9969,
+      "step": 5830
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.795622354607419e-05,
+      "loss": 2.0213,
+      "step": 5840
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.79486503174729e-05,
+      "loss": 1.9825,
+      "step": 5850
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.7941063580147637e-05,
+      "loss": 2.0005,
+      "step": 5860
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.793346333776539e-05,
+      "loss": 2.0043,
+      "step": 5870
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.7925849593999656e-05,
+      "loss": 1.9907,
+      "step": 5880
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.791822235253046e-05,
+      "loss": 1.9813,
+      "step": 5890
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.7910581617044375e-05,
+      "loss": 1.987,
+      "step": 5900
+    },
+    {
+      "epoch": 0.26,
+      "eval_loss": 1.9848577976226807,
+      "eval_runtime": 13.9355,
+      "eval_samples_per_second": 293.926,
+      "eval_steps_per_second": 18.37,
+      "step": 5900
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.790292739123448e-05,
+      "loss": 2.0303,
+      "step": 5910
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.789525967880037e-05,
+      "loss": 1.9622,
+      "step": 5920
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.788757848344816e-05,
+      "loss": 1.9876,
+      "step": 5930
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.787988380889048e-05,
+      "loss": 1.9844,
+      "step": 5940
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.7872175658846495e-05,
+      "loss": 1.9737,
+      "step": 5950
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.7864454037041864e-05,
+      "loss": 1.992,
+      "step": 5960
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.785671894720876e-05,
+      "loss": 2.0209,
+      "step": 5970
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.784897039308587e-05,
+      "loss": 1.9725,
+      "step": 5980
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.7841208378418386e-05,
+      "loss": 1.9777,
+      "step": 5990
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.783343290695801e-05,
+      "loss": 2.0128,
+      "step": 6000
+    },
+    {
+      "epoch": 0.26,
+      "eval_loss": 1.9819371700286865,
+      "eval_runtime": 12.081,
+      "eval_samples_per_second": 339.046,
+      "eval_steps_per_second": 21.19,
+      "step": 6000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.7825643982462926e-05,
+      "loss": 1.9872,
+      "step": 6010
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.781784160869786e-05,
+      "loss": 2.0036,
+      "step": 6020
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.7810025789434014e-05,
+      "loss": 1.984,
+      "step": 6030
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.780219652844909e-05,
+      "loss": 1.9859,
+      "step": 6040
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.77943538295273e-05,
+      "loss": 2.0032,
+      "step": 6050
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.778649769645932e-05,
+      "loss": 2.0241,
+      "step": 6060
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.7778628133042346e-05,
+      "loss": 2.0075,
+      "step": 6070
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.7770745143080076e-05,
+      "loss": 1.9937,
+      "step": 6080
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.776284873038266e-05,
+      "loss": 2.0228,
+      "step": 6090
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.775493889876677e-05,
+      "loss": 1.9932,
+      "step": 6100
+    },
+    {
+      "epoch": 0.27,
+      "eval_loss": 1.9792561531066895,
+      "eval_runtime": 15.4668,
+      "eval_samples_per_second": 264.825,
+      "eval_steps_per_second": 16.552,
+      "step": 6100
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.774701565205553e-05,
+      "loss": 1.9778,
+      "step": 6110
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.77390789940786e-05,
+      "loss": 2.0101,
+      "step": 6120
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.773112892867207e-05,
+      "loss": 1.9633,
+      "step": 6130
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.7723165459678554e-05,
+      "loss": 1.9755,
+      "step": 6140
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.771518859094709e-05,
+      "loss": 2.0302,
+      "step": 6150
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.770719832633324e-05,
+      "loss": 1.961,
+      "step": 6160
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.7699194669699026e-05,
+      "loss": 1.9772,
+      "step": 6170
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.7691177624912934e-05,
+      "loss": 1.9707,
+      "step": 6180
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.768314719584995e-05,
+      "loss": 1.9767,
+      "step": 6190
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.767510338639148e-05,
+      "loss": 1.9874,
+      "step": 6200
+    },
+    {
+      "epoch": 0.27,
+      "eval_loss": 1.9786502122879028,
+      "eval_runtime": 13.4256,
+      "eval_samples_per_second": 305.088,
+      "eval_steps_per_second": 19.068,
+      "step": 6200
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.7667046200425445e-05,
+      "loss": 1.9606,
+      "step": 6210
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.76589756418462e-05,
+      "loss": 1.9976,
+      "step": 6220
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.765089171455459e-05,
+      "loss": 2.0065,
+      "step": 6230
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.764279442245789e-05,
+      "loss": 2.0063,
+      "step": 6240
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.763468376946987e-05,
+      "loss": 1.9764,
+      "step": 6250
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.762655975951073e-05,
+      "loss": 2.0229,
+      "step": 6260
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.761842239650713e-05,
+      "loss": 1.9482,
+      "step": 6270
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.761027168439219e-05,
+      "loss": 2.0053,
+      "step": 6280
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.7602107627105506e-05,
+      "loss": 2.0104,
+      "step": 6290
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.759393022859308e-05,
+      "loss": 1.9935,
+      "step": 6300
+    },
+    {
+      "epoch": 0.28,
+      "eval_loss": 1.9752275943756104,
+      "eval_runtime": 13.0612,
+      "eval_samples_per_second": 313.6,
+      "eval_steps_per_second": 19.6,
+      "step": 6300
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.758573949280737e-05,
+      "loss": 1.9944,
+      "step": 6310
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.7577535423707325e-05,
+      "loss": 1.977,
+      "step": 6320
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.7569318025258286e-05,
+      "loss": 1.9996,
+      "step": 6330
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.7561087301432054e-05,
+      "loss": 1.9515,
+      "step": 6340
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.755284325620688e-05,
+      "loss": 1.9979,
+      "step": 6350
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.754458589356746e-05,
+      "loss": 1.9294,
+      "step": 6360
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.753631521750489e-05,
+      "loss": 1.9437,
+      "step": 6370
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.7528031232016744e-05,
+      "loss": 1.944,
+      "step": 6380
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.751973394110701e-05,
+      "loss": 1.9716,
+      "step": 6390
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.751142334878609e-05,
+      "loss": 1.9515,
+      "step": 6400
+    },
+    {
+      "epoch": 0.28,
+      "eval_loss": 1.972642183303833,
+      "eval_runtime": 13.9092,
+      "eval_samples_per_second": 294.481,
+      "eval_steps_per_second": 18.405,
+      "step": 6400
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.7503099459070864e-05,
+      "loss": 1.9862,
+      "step": 6410
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.749476227598457e-05,
+      "loss": 1.9563,
+      "step": 6420
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.748641180355694e-05,
+      "loss": 2.0096,
+      "step": 6430
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.747804804582409e-05,
+      "loss": 2.0074,
+      "step": 6440
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.7469671006828546e-05,
+      "loss": 1.9559,
+      "step": 6450
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.7461280690619305e-05,
+      "loss": 1.9869,
+      "step": 6460
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.745287710125173e-05,
+      "loss": 1.9906,
+      "step": 6470
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.7444460242787614e-05,
+      "loss": 2.0531,
+      "step": 6480
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.7436030119295184e-05,
+      "loss": 1.9787,
+      "step": 6490
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.742758673484905e-05,
+      "loss": 2.0037,
+      "step": 6500
+    },
+    {
+      "epoch": 0.28,
+      "eval_loss": 1.9702032804489136,
+      "eval_runtime": 12.617,
+      "eval_samples_per_second": 324.641,
+      "eval_steps_per_second": 20.29,
+      "step": 6500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.7419130093530255e-05,
+      "loss": 2.0547,
+      "step": 6510
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.7410660199426236e-05,
+      "loss": 2.052,
+      "step": 6520
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.740217705663083e-05,
+      "loss": 1.9391,
+      "step": 6530
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.739368066924431e-05,
+      "loss": 1.9846,
+      "step": 6540
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.7385171041373285e-05,
+      "loss": 2.0112,
+      "step": 6550
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.7376648177130846e-05,
+      "loss": 1.9748,
+      "step": 6560
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.736811208063642e-05,
+      "loss": 1.9662,
+      "step": 6570
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.7359562756015864e-05,
+      "loss": 1.9933,
+      "step": 6580
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.735100020740141e-05,
+      "loss": 2.0313,
+      "step": 6590
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.734242443893168e-05,
+      "loss": 1.9313,
+      "step": 6600
+    },
+    {
+      "epoch": 0.29,
+      "eval_loss": 1.9688479900360107,
+      "eval_runtime": 12.9347,
+      "eval_samples_per_second": 316.667,
+      "eval_steps_per_second": 19.792,
+      "step": 6600
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.73338354547517e-05,
+      "loss": 1.9515,
+      "step": 6610
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.732523325901288e-05,
+      "loss": 1.954,
+      "step": 6620
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.731661785587301e-05,
+      "loss": 2.0307,
+      "step": 6630
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.730798924949626e-05,
+      "loss": 2.001,
+      "step": 6640
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.729934744405319e-05,
+      "loss": 2.0087,
+      "step": 6650
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.729069244372075e-05,
+      "loss": 2.0286,
+      "step": 6660
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.728202425268224e-05,
+      "loss": 2.0152,
+      "step": 6670
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.727334287512735e-05,
+      "loss": 1.9232,
+      "step": 6680
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.7264648315252156e-05,
+      "loss": 1.976,
+      "step": 6690
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.725594057725909e-05,
+      "loss": 1.968,
+      "step": 6700
+    },
+    {
+      "epoch": 0.29,
+      "eval_loss": 1.9677083492279053,
+      "eval_runtime": 12.2837,
+      "eval_samples_per_second": 333.451,
+      "eval_steps_per_second": 20.841,
+      "step": 6700
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.724721966535695e-05,
+      "loss": 1.9615,
+      "step": 6710
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.723848558376092e-05,
+      "loss": 1.9948,
+      "step": 6720
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.722973833669252e-05,
+      "loss": 1.9942,
+      "step": 6730
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.7220977928379684e-05,
+      "loss": 2.0166,
+      "step": 6740
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.721220436305664e-05,
+      "loss": 1.9345,
+      "step": 6750
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.720341764496404e-05,
+      "loss": 2.0107,
+      "step": 6760
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.719461777834883e-05,
+      "loss": 1.9749,
+      "step": 6770
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.7185804767464375e-05,
+      "loss": 2.0009,
+      "step": 6780
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.717697861657035e-05,
+      "loss": 2.0079,
+      "step": 6790
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.7168139329932796e-05,
+      "loss": 1.9954,
+      "step": 6800
+    },
+    {
+      "epoch": 0.3,
+      "eval_loss": 1.9654452800750732,
+      "eval_runtime": 12.8097,
+      "eval_samples_per_second": 319.759,
+      "eval_steps_per_second": 19.985,
+      "step": 6800
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.71592869118241e-05,
+      "loss": 2.0057,
+      "step": 6810
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.7150421366523e-05,
+      "loss": 1.9505,
+      "step": 6820
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.7141542698314585e-05,
+      "loss": 2.0055,
+      "step": 6830
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.713265091149025e-05,
+      "loss": 1.9874,
+      "step": 6840
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.71237460103478e-05,
+      "loss": 1.9679,
+      "step": 6850
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.711482799919129e-05,
+      "loss": 1.9643,
+      "step": 6860
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.710589688233119e-05,
+      "loss": 2.0399,
+      "step": 6870
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.709695266408426e-05,
+      "loss": 1.9797,
+      "step": 6880
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.708799534877363e-05,
+      "loss": 1.9537,
+      "step": 6890
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.707902494072871e-05,
+      "loss": 1.9182,
+      "step": 6900
+    },
+    {
+      "epoch": 0.3,
+      "eval_loss": 1.9642119407653809,
+      "eval_runtime": 13.2593,
+      "eval_samples_per_second": 308.916,
+      "eval_steps_per_second": 19.307,
+      "step": 6900
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.707004144428526e-05,
+      "loss": 2.007,
+      "step": 6910
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.706104486378539e-05,
+      "loss": 1.9933,
+      "step": 6920
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.7052035203577516e-05,
+      "loss": 2.013,
+      "step": 6930
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.704301246801636e-05,
+      "loss": 1.9395,
+      "step": 6940
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.703397666146299e-05,
+      "loss": 1.9651,
+      "step": 6950
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.7024927788284765e-05,
+      "loss": 1.9879,
+      "step": 6960
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.7015865852855406e-05,
+      "loss": 1.9733,
+      "step": 6970
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.700679085955488e-05,
+      "loss": 2.0089,
+      "step": 6980
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.699770281276952e-05,
+      "loss": 2.0274,
+      "step": 6990
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.6988601716891954e-05,
+      "loss": 1.9599,
+      "step": 7000
+    },
+    {
+      "epoch": 0.31,
+      "eval_loss": 1.9618194103240967,
+      "eval_runtime": 13.1445,
+      "eval_samples_per_second": 311.613,
+      "eval_steps_per_second": 19.476,
+      "step": 7000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.69794875763211e-05,
+      "loss": 1.9609,
+      "step": 7010
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.6970360395462204e-05,
+      "loss": 1.9311,
+      "step": 7020
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.69612201787268e-05,
+      "loss": 2.0024,
+      "step": 7030
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.695206693053273e-05,
+      "loss": 1.9428,
+      "step": 7040
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.694290065530414e-05,
+      "loss": 1.9586,
+      "step": 7050
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.6933721357471455e-05,
+      "loss": 1.967,
+      "step": 7060
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.692452904147141e-05,
+      "loss": 2.0423,
+      "step": 7070
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.6915323711747023e-05,
+      "loss": 1.9693,
+      "step": 7080
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.69061053727476e-05,
+      "loss": 2.001,
+      "step": 7090
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.689687402892876e-05,
+      "loss": 1.9601,
+      "step": 7100
+    },
+    {
+      "epoch": 0.31,
+      "eval_loss": 1.959427833557129,
+      "eval_runtime": 12.6915,
+      "eval_samples_per_second": 322.737,
+      "eval_steps_per_second": 20.171,
+      "step": 7100
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.688762968475237e-05,
+      "loss": 1.9979,
+      "step": 7110
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.687837234468661e-05,
+      "loss": 1.9676,
+      "step": 7120
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.686910201320592e-05,
+      "loss": 2.0003,
+      "step": 7130
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.685981869479104e-05,
+      "loss": 1.9988,
+      "step": 7140
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.685052239392897e-05,
+      "loss": 2.0026,
+      "step": 7150
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.6841213115113e-05,
+      "loss": 1.9073,
+      "step": 7160
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.683189086284268e-05,
+      "loss": 1.9597,
+      "step": 7170
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.682255564162382e-05,
+      "loss": 1.9903,
+      "step": 7180
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.6813207455968534e-05,
+      "loss": 2.0011,
+      "step": 7190
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.680384631039519e-05,
+      "loss": 1.9978,
+      "step": 7200
+    },
+    {
+      "epoch": 0.32,
+      "eval_loss": 1.9591736793518066,
+      "eval_runtime": 14.3637,
+      "eval_samples_per_second": 285.163,
+      "eval_steps_per_second": 17.823,
+      "step": 7200
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.679447220942838e-05,
+      "loss": 1.976,
+      "step": 7210
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.6785085157599016e-05,
+      "loss": 1.9745,
+      "step": 7220
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.677568515944424e-05,
+      "loss": 1.968,
+      "step": 7230
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.676627221950743e-05,
+      "loss": 1.993,
+      "step": 7240
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.675684634233828e-05,
+      "loss": 1.979,
+      "step": 7250
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.674740753249268e-05,
+      "loss": 2.0276,
+      "step": 7260
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.673795579453281e-05,
+      "loss": 1.9836,
+      "step": 7270
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.672849113302705e-05,
+      "loss": 1.9735,
+      "step": 7280
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.6719013552550084e-05,
+      "loss": 1.9811,
+      "step": 7290
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.67095230576828e-05,
+      "loss": 1.9934,
+      "step": 7300
+    },
+    {
+      "epoch": 0.32,
+      "eval_loss": 1.9573659896850586,
+      "eval_runtime": 13.2709,
+      "eval_samples_per_second": 308.645,
+      "eval_steps_per_second": 19.29,
+      "step": 7300
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.6700019653012354e-05,
+      "loss": 2.0308,
+      "step": 7310
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.669050334313213e-05,
+      "loss": 1.9438,
+      "step": 7320
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.668097413264173e-05,
+      "loss": 1.9725,
+      "step": 7330
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.6671432026147035e-05,
+      "loss": 1.9757,
+      "step": 7340
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.6661877028260107e-05,
+      "loss": 1.9539,
+      "step": 7350
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.665230914359929e-05,
+      "loss": 1.998,
+      "step": 7360
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.6642728376789124e-05,
+      "loss": 1.9701,
+      "step": 7370
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.663313473246038e-05,
+      "loss": 1.9892,
+      "step": 7380
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.662352821525007e-05,
+      "loss": 2.0117,
+      "step": 7390
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.661390882980141e-05,
+      "loss": 2.0,
+      "step": 7400
+    },
+    {
+      "epoch": 0.32,
+      "eval_loss": 1.9556201696395874,
+      "eval_runtime": 13.8205,
+      "eval_samples_per_second": 296.372,
+      "eval_steps_per_second": 18.523,
+      "step": 7400
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.660427658076384e-05,
+      "loss": 2.0104,
+      "step": 7410
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.6594631472793035e-05,
+      "loss": 1.9395,
+      "step": 7420
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.6584973510550844e-05,
+      "loss": 1.9631,
+      "step": 7430
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.657530269870536e-05,
+      "loss": 1.9215,
+      "step": 7440
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.6565619041930894e-05,
+      "loss": 1.9726,
+      "step": 7450
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.655592254490795e-05,
+      "loss": 1.9995,
+      "step": 7460
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.6546213212323235e-05,
+      "loss": 2.0179,
+      "step": 7470
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.6536491048869666e-05,
+      "loss": 1.9594,
+      "step": 7480
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.6526756059246365e-05,
+      "loss": 1.9821,
+      "step": 7490
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.651700824815865e-05,
+      "loss": 1.9719,
+      "step": 7500
+    },
+    {
+      "epoch": 0.33,
+      "eval_loss": 1.9540364742279053,
+      "eval_runtime": 14.3302,
+      "eval_samples_per_second": 285.83,
+      "eval_steps_per_second": 17.864,
+      "step": 7500
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.650724762031803e-05,
+      "loss": 1.9848,
+      "step": 7510
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.6497474180442224e-05,
+      "loss": 1.9471,
+      "step": 7520
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.648768793325513e-05,
+      "loss": 1.9697,
+      "step": 7530
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.6477888883486844e-05,
+      "loss": 1.9522,
+      "step": 7540
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.646807703587365e-05,
+      "loss": 1.9738,
+      "step": 7550
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.645825239515799e-05,
+      "loss": 1.9944,
+      "step": 7560
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.644841496608855e-05,
+      "loss": 2.0082,
+      "step": 7570
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.643856475342014e-05,
+      "loss": 2.0052,
+      "step": 7580
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.642870176191377e-05,
+      "loss": 1.9821,
+      "step": 7590
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.641882599633663e-05,
+      "loss": 1.991,
+      "step": 7600
+    },
+    {
+      "epoch": 0.33,
+      "eval_loss": 1.951909065246582,
+      "eval_runtime": 12.5953,
+      "eval_samples_per_second": 325.201,
+      "eval_steps_per_second": 20.325,
+      "step": 7600
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.6408937461462095e-05,
+      "loss": 2.0207,
+      "step": 7610
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.639903616206967e-05,
+      "loss": 1.9723,
+      "step": 7620
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.638912210294508e-05,
+      "loss": 1.9881,
+      "step": 7630
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.6379195288880184e-05,
+      "loss": 1.955,
+      "step": 7640
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.636925572467301e-05,
+      "loss": 1.9749,
+      "step": 7650
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.635930341512778e-05,
+      "loss": 1.9815,
+      "step": 7660
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.6349338365054825e-05,
+      "loss": 1.9807,
+      "step": 7670
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.633936057927067e-05,
+      "loss": 1.9934,
+      "step": 7680
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.632937006259799e-05,
+      "loss": 1.9536,
+      "step": 7690
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.631936681986561e-05,
+      "loss": 1.9894,
+      "step": 7700
+    },
+    {
+      "epoch": 0.34,
+      "eval_loss": 1.9506793022155762,
+      "eval_runtime": 13.5546,
+      "eval_samples_per_second": 302.184,
+      "eval_steps_per_second": 18.887,
+      "step": 7700
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.63093508559085e-05,
+      "loss": 1.9473,
+      "step": 7710
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.6299322175567783e-05,
+      "loss": 1.98,
+      "step": 7720
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.628928078369074e-05,
+      "loss": 1.9744,
+      "step": 7730
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.6279226685130784e-05,
+      "loss": 1.9851,
+      "step": 7740
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.626915988474746e-05,
+      "loss": 1.9756,
+      "step": 7750
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.625908038740647e-05,
+      "loss": 2.035,
+      "step": 7760
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.6248988197979664e-05,
+      "loss": 1.9488,
+      "step": 7770
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.623888332134497e-05,
+      "loss": 1.9579,
+      "step": 7780
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.622876576238652e-05,
+      "loss": 1.9861,
+      "step": 7790
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.6218635525994545e-05,
+      "loss": 1.9895,
+      "step": 7800
+    },
+    {
+      "epoch": 0.34,
+      "eval_loss": 1.9489729404449463,
+      "eval_runtime": 12.8919,
+      "eval_samples_per_second": 317.719,
+      "eval_steps_per_second": 19.857,
+      "step": 7800
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.620849261706537e-05,
+      "loss": 2.0395,
+      "step": 7810
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.6198337040501505e-05,
+      "loss": 1.958,
+      "step": 7820
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.618816880121155e-05,
+      "loss": 1.9912,
+      "step": 7830
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.617798790411023e-05,
+      "loss": 1.9098,
+      "step": 7840
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.616779435411838e-05,
+      "loss": 1.9905,
+      "step": 7850
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.615758815616297e-05,
+      "loss": 1.9593,
+      "step": 7860
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.614736931517707e-05,
+      "loss": 1.9738,
+      "step": 7870
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.613713783609985e-05,
+      "loss": 1.9318,
+      "step": 7880
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.612689372387662e-05,
+      "loss": 1.9924,
+      "step": 7890
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.6116636983458766e-05,
+      "loss": 1.97,
+      "step": 7900
+    },
+    {
+      "epoch": 0.35,
+      "eval_loss": 1.947704553604126,
+      "eval_runtime": 12.113,
+      "eval_samples_per_second": 338.148,
+      "eval_steps_per_second": 21.134,
+      "step": 7900
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.61063676198038e-05,
+      "loss": 1.9341,
+      "step": 7910
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.609608563787532e-05,
+      "loss": 2.0017,
+      "step": 7920
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.608579104264301e-05,
+      "loss": 1.9553,
+      "step": 7930
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.607548383908269e-05,
+      "loss": 1.9659,
+      "step": 7940
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.6065164032176255e-05,
+      "loss": 2.0098,
+      "step": 7950
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.605483162691168e-05,
+      "loss": 1.9583,
+      "step": 7960
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.6044486628283026e-05,
+      "loss": 1.8988,
+      "step": 7970
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.6034129041290476e-05,
+      "loss": 1.9599,
+      "step": 7980
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.6023758870940265e-05,
+      "loss": 1.9926,
+      "step": 7990
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.601337612224472e-05,
+      "loss": 1.9334,
+      "step": 8000
+    },
+    {
+      "epoch": 0.35,
+      "eval_loss": 1.9458627700805664,
+      "eval_runtime": 13.6514,
+      "eval_samples_per_second": 300.043,
+      "eval_steps_per_second": 18.753,
+      "step": 8000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.600298080022224e-05,
+      "loss": 1.9365,
+      "step": 8010
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.599257290989733e-05,
+      "loss": 2.0154,
+      "step": 8020
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.5982152456300514e-05,
+      "loss": 1.9472,
+      "step": 8030
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.5971719444468454e-05,
+      "loss": 1.9347,
+      "step": 8040
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.5961273879443845e-05,
+      "loss": 1.9706,
+      "step": 8050
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.595081576627544e-05,
+      "loss": 1.8983,
+      "step": 8060
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.594034511001809e-05,
+      "loss": 1.9517,
+      "step": 8070
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.5929861915732674e-05,
+      "loss": 2.0011,
+      "step": 8080
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.591936618848617e-05,
+      "loss": 1.9648,
+      "step": 8090
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.590885793335156e-05,
+      "loss": 1.9856,
+      "step": 8100
+    },
+    {
+      "epoch": 0.35,
+      "eval_loss": 1.9449169635772705,
+      "eval_runtime": 12.4323,
+      "eval_samples_per_second": 329.464,
+      "eval_steps_per_second": 20.591,
+      "step": 8100
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.589833715540794e-05,
+      "loss": 1.9837,
+      "step": 8110
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.588780385974042e-05,
+      "loss": 1.996,
+      "step": 8120
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.5877258051440175e-05,
+      "loss": 1.9771,
+      "step": 8130
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.5866699735604424e-05,
+      "loss": 1.9989,
+      "step": 8140
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.5856128917336437e-05,
+      "loss": 1.99,
+      "step": 8150
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.584554560174552e-05,
+      "loss": 2.0055,
+      "step": 8160
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.5834949793947004e-05,
+      "loss": 1.9763,
+      "step": 8170
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.5824341499062304e-05,
+      "loss": 1.9356,
+      "step": 8180
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.581372072221882e-05,
+      "loss": 1.9056,
+      "step": 8190
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.580308746855002e-05,
+      "loss": 1.9748,
+      "step": 8200
+    },
+    {
+      "epoch": 0.36,
+      "eval_loss": 1.9407728910446167,
+      "eval_runtime": 14.8862,
+      "eval_samples_per_second": 275.154,
+      "eval_steps_per_second": 17.197,
+      "step": 8200
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.5792441743195385e-05,
+      "loss": 1.9251,
+      "step": 8210
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.5781783551300435e-05,
+      "loss": 1.9767,
+      "step": 8220
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.57711128980167e-05,
+      "loss": 1.9116,
+      "step": 8230
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.576042978850174e-05,
+      "loss": 1.9413,
+      "step": 8240
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.574973422791916e-05,
+      "loss": 1.9468,
+      "step": 8250
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.573902622143852e-05,
+      "loss": 1.9562,
+      "step": 8260
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.572830577423549e-05,
+      "loss": 1.9744,
+      "step": 8270
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.571757289149165e-05,
+      "loss": 1.9422,
+      "step": 8280
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.570682757839467e-05,
+      "loss": 1.9565,
+      "step": 8290
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.569606984013819e-05,
+      "loss": 2.0056,
+      "step": 8300
+    },
+    {
+      "epoch": 0.36,
+      "eval_loss": 1.9395599365234375,
+      "eval_runtime": 11.71,
+      "eval_samples_per_second": 349.786,
+      "eval_steps_per_second": 21.862,
+      "step": 8300
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.568529968192187e-05,
+      "loss": 1.9783,
+      "step": 8310
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.5674517108951365e-05,
+      "loss": 1.9335,
+      "step": 8320
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.5663722126438316e-05,
+      "loss": 1.9209,
+      "step": 8330
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.565291473960041e-05,
+      "loss": 1.9587,
+      "step": 8340
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.5642094953661274e-05,
+      "loss": 2.0141,
+      "step": 8350
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.563126277385056e-05,
+      "loss": 1.9702,
+      "step": 8360
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.562041820540391e-05,
+      "loss": 2.0086,
+      "step": 8370
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.560956125356293e-05,
+      "loss": 1.9322,
+      "step": 8380
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.5598691923575244e-05,
+      "loss": 1.9788,
+      "step": 8390
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.558781022069443e-05,
+      "loss": 1.982,
+      "step": 8400
+    },
+    {
+      "epoch": 0.37,
+      "eval_loss": 1.9378072023391724,
+      "eval_runtime": 13.6976,
+      "eval_samples_per_second": 299.029,
+      "eval_steps_per_second": 18.689,
+      "step": 8400
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.557691615018008e-05,
+      "loss": 2.0186,
+      "step": 8410
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.556600971729771e-05,
+      "loss": 1.9675,
+      "step": 8420
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.5555090927318876e-05,
+      "loss": 1.9675,
+      "step": 8430
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.554415978552106e-05,
+      "loss": 2.0175,
+      "step": 8440
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.5533216297187726e-05,
+      "loss": 1.9642,
+      "step": 8450
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.5522260467608326e-05,
+      "loss": 1.9741,
+      "step": 8460
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.5511292302078244e-05,
+      "loss": 1.9557,
+      "step": 8470
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.550031180589884e-05,
+      "loss": 1.9811,
+      "step": 8480
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.5489318984377456e-05,
+      "loss": 1.9578,
+      "step": 8490
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.547831384282734e-05,
+      "loss": 1.9934,
+      "step": 8500
+    },
+    {
+      "epoch": 0.37,
+      "eval_loss": 1.9361932277679443,
+      "eval_runtime": 12.095,
+      "eval_samples_per_second": 338.652,
+      "eval_steps_per_second": 21.166,
+      "step": 8500
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.546729638656777e-05,
+      "loss": 1.9608,
+      "step": 8510
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.545626662092389e-05,
+      "loss": 1.966,
+      "step": 8520
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.544522455122686e-05,
+      "loss": 1.9437,
+      "step": 8530
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.5434170182813755e-05,
+      "loss": 1.9392,
+      "step": 8540
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.54231035210276e-05,
+      "loss": 1.9426,
+      "step": 8550
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.541202457121737e-05,
+      "loss": 1.9498,
+      "step": 8560
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.540093333873798e-05,
+      "loss": 1.9778,
+      "step": 8570
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.5389829828950266e-05,
+      "loss": 1.9281,
+      "step": 8580
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.5378714047221007e-05,
+      "loss": 1.99,
+      "step": 8590
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.5367585998922916e-05,
+      "loss": 1.9328,
+      "step": 8600
+    },
+    {
+      "epoch": 0.38,
+      "eval_loss": 1.9360263347625732,
+      "eval_runtime": 11.7734,
+      "eval_samples_per_second": 347.904,
+      "eval_steps_per_second": 21.744,
+      "step": 8600
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.535644568943464e-05,
+      "loss": 1.9399,
+      "step": 8610
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.5345293124140736e-05,
+      "loss": 1.9359,
+      "step": 8620
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.533412830843169e-05,
+      "loss": 1.9176,
+      "step": 8630
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.5322951247703935e-05,
+      "loss": 1.9591,
+      "step": 8640
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.5311761947359786e-05,
+      "loss": 1.964,
+      "step": 8650
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.530056041280748e-05,
+      "loss": 1.9774,
+      "step": 8660
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.5289346649461196e-05,
+      "loss": 1.9532,
+      "step": 8670
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.527812066274099e-05,
+      "loss": 1.9559,
+      "step": 8680
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.5266882458072845e-05,
+      "loss": 1.9101,
+      "step": 8690
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.525563204088864e-05,
+      "loss": 2.0133,
+      "step": 8700
+    },
+    {
+      "epoch": 0.38,
+      "eval_loss": 1.9347482919692993,
+      "eval_runtime": 11.5929,
+      "eval_samples_per_second": 353.32,
+      "eval_steps_per_second": 22.082,
+      "step": 8700
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.524436941662616e-05,
+      "loss": 1.9481,
+      "step": 8710
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.52330945907291e-05,
+      "loss": 2.0243,
+      "step": 8720
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.522180756864705e-05,
+      "loss": 2.0271,
+      "step": 8730
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.521050835583546e-05,
+      "loss": 1.9358,
+      "step": 8740
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.519919695775572e-05,
+      "loss": 2.0078,
+      "step": 8750
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.51878733798751e-05,
+      "loss": 1.9632,
+      "step": 8760
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.517653762766673e-05,
+      "loss": 1.9393,
+      "step": 8770
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.516518970660965e-05,
+      "loss": 1.9215,
+      "step": 8780
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.515382962218877e-05,
+      "loss": 1.9502,
+      "step": 8790
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.514245737989489e-05,
+      "loss": 1.9523,
+      "step": 8800
+    },
+    {
+      "epoch": 0.39,
+      "eval_loss": 1.9324275255203247,
+      "eval_runtime": 11.8045,
+      "eval_samples_per_second": 346.986,
+      "eval_steps_per_second": 21.687,
+      "step": 8800
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.513107298522467e-05,
+      "loss": 1.9323,
+      "step": 8810
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.5119676443680657e-05,
+      "loss": 1.9541,
+      "step": 8820
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.5108267760771266e-05,
+      "loss": 1.9716,
+      "step": 8830
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.5096846942010775e-05,
+      "loss": 1.9488,
+      "step": 8840
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.5085413992919344e-05,
+      "loss": 1.9172,
+      "step": 8850
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.507396891902296e-05,
+      "loss": 1.9546,
+      "step": 8860
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.506251172585352e-05,
+      "loss": 1.9807,
+      "step": 8870
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.505104241894875e-05,
+      "loss": 1.9575,
+      "step": 8880
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.503956100385221e-05,
+      "loss": 1.963,
+      "step": 8890
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.5028067486113366e-05,
+      "loss": 1.954,
+      "step": 8900
+    },
+    {
+      "epoch": 0.39,
+      "eval_loss": 1.9312846660614014,
+      "eval_runtime": 11.7288,
+      "eval_samples_per_second": 349.225,
+      "eval_steps_per_second": 21.827,
+      "step": 8900
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.5016561871287496e-05,
+      "loss": 1.9755,
+      "step": 8910
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.500504416493572e-05,
+      "loss": 1.9325,
+      "step": 8920
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.499351437262503e-05,
+      "loss": 2.0681,
+      "step": 8930
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.4981972499928244e-05,
+      "loss": 1.9759,
+      "step": 8940
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.497041855242401e-05,
+      "loss": 1.9578,
+      "step": 8950
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.495885253569684e-05,
+      "loss": 1.9771,
+      "step": 8960
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.4947274455337046e-05,
+      "loss": 1.9007,
+      "step": 8970
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.49356843169408e-05,
+      "loss": 1.9742,
+      "step": 8980
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.492408212611009e-05,
+      "loss": 1.9316,
+      "step": 8990
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.4912467888452705e-05,
+      "loss": 1.9556,
+      "step": 9000
+    },
+    {
+      "epoch": 0.39,
+      "eval_loss": 1.9298361539840698,
+      "eval_runtime": 11.9118,
+      "eval_samples_per_second": 343.861,
+      "eval_steps_per_second": 21.491,
+      "step": 9000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.4900841609582304e-05,
+      "loss": 1.9705,
+      "step": 9010
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.488920329511833e-05,
+      "loss": 1.9575,
+      "step": 9020
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.487755295068606e-05,
+      "loss": 1.9734,
+      "step": 9030
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.486589058191659e-05,
+      "loss": 2.0,
+      "step": 9040
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.4854216194446806e-05,
+      "loss": 1.9733,
+      "step": 9050
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.484252979391942e-05,
+      "loss": 1.9879,
+      "step": 9060
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.4830831385982955e-05,
+      "loss": 1.9072,
+      "step": 9070
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.4819120976291726e-05,
+      "loss": 1.9627,
+      "step": 9080
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.4807398570505836e-05,
+      "loss": 1.9815,
+      "step": 9090
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.4795664174291214e-05,
+      "loss": 1.975,
+      "step": 9100
+    },
+    {
+      "epoch": 0.4,
+      "eval_loss": 1.9286532402038574,
+      "eval_runtime": 11.9769,
+      "eval_samples_per_second": 341.991,
+      "eval_steps_per_second": 21.374,
+      "step": 9100
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.478391779331958e-05,
+      "loss": 1.935,
+      "step": 9110
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.477215943326843e-05,
+      "loss": 1.9991,
+      "step": 9120
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.476038909982106e-05,
+      "loss": 1.958,
+      "step": 9130
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.474860679866656e-05,
+      "loss": 1.9126,
+      "step": 9140
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.47368125354998e-05,
+      "loss": 1.9441,
+      "step": 9150
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.4725006316021404e-05,
+      "loss": 1.9427,
+      "step": 9160
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.471318814593783e-05,
+      "loss": 1.9856,
+      "step": 9170
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.4701358030961266e-05,
+      "loss": 1.9876,
+      "step": 9180
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.468951597680969e-05,
+      "loss": 1.9421,
+      "step": 9190
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.467766198920686e-05,
+      "loss": 1.9562,
+      "step": 9200
+    },
+    {
+      "epoch": 0.4,
+      "eval_loss": 1.9275963306427002,
+      "eval_runtime": 11.7257,
+      "eval_samples_per_second": 349.317,
+      "eval_steps_per_second": 21.832,
+      "step": 9200
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.466579607388229e-05,
+      "loss": 2.0232,
+      "step": 9210
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.4653918236571245e-05,
+      "loss": 1.8672,
+      "step": 9220
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.464202848301479e-05,
+      "loss": 1.9394,
+      "step": 9230
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.463012681895972e-05,
+      "loss": 1.9189,
+      "step": 9240
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.461821325015859e-05,
+      "loss": 1.9301,
+      "step": 9250
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.4606287782369724e-05,
+      "loss": 1.9574,
+      "step": 9260
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.459435042135718e-05,
+      "loss": 1.9584,
+      "step": 9270
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.458240117289077e-05,
+      "loss": 1.9699,
+      "step": 9280
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.457044004274607e-05,
+      "loss": 1.976,
+      "step": 9290
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.455846703670436e-05,
+      "loss": 1.9704,
+      "step": 9300
+    },
+    {
+      "epoch": 0.41,
+      "eval_loss": 1.9267959594726562,
+      "eval_runtime": 11.5085,
+      "eval_samples_per_second": 355.912,
+      "eval_steps_per_second": 22.244,
+      "step": 9300
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.45464821605527e-05,
+      "loss": 1.9217,
+      "step": 9310
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.4534485420083866e-05,
+      "loss": 1.9389,
+      "step": 9320
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.4522476821096364e-05,
+      "loss": 1.9367,
+      "step": 9330
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.4510456369394455e-05,
+      "loss": 1.9851,
+      "step": 9340
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.44984240707881e-05,
+      "loss": 1.9447,
+      "step": 9350
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.448637993109301e-05,
+      "loss": 1.9024,
+      "step": 9360
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.44743239561306e-05,
+      "loss": 1.9409,
+      "step": 9370
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.4462256151728024e-05,
+      "loss": 1.9617,
+      "step": 9380
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.4450176523718144e-05,
+      "loss": 1.9452,
+      "step": 9390
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.443808507793953e-05,
+      "loss": 1.9769,
+      "step": 9400
+    },
+    {
+      "epoch": 0.41,
+      "eval_loss": 1.9254794120788574,
+      "eval_runtime": 11.6603,
+      "eval_samples_per_second": 351.277,
+      "eval_steps_per_second": 21.955,
+      "step": 9400
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.442598182023648e-05,
+      "loss": 1.9595,
+      "step": 9410
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.441386675645899e-05,
+      "loss": 1.9119,
+      "step": 9420
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.440173989246276e-05,
+      "loss": 1.9025,
+      "step": 9430
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.438960123410921e-05,
+      "loss": 1.9572,
+      "step": 9440
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.437745078726543e-05,
+      "loss": 1.9962,
+      "step": 9450
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.4365288557804243e-05,
+      "loss": 1.9908,
+      "step": 9460
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.435311455160415e-05,
+      "loss": 1.9532,
+      "step": 9470
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.434092877454934e-05,
+      "loss": 1.9798,
+      "step": 9480
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.43287312325297e-05,
+      "loss": 1.894,
+      "step": 9490
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.43165219314408e-05,
+      "loss": 1.9114,
+      "step": 9500
+    },
+    {
+      "epoch": 0.42,
+      "eval_loss": 1.923863410949707,
+      "eval_runtime": 12.1484,
+      "eval_samples_per_second": 337.164,
+      "eval_steps_per_second": 21.073,
+      "step": 9500
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.43043008771839e-05,
+      "loss": 2.0034,
+      "step": 9510
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.429206807566592e-05,
+      "loss": 1.9252,
+      "step": 9520
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.4279823532799476e-05,
+      "loss": 1.9489,
+      "step": 9530
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.4267567254502865e-05,
+      "loss": 1.9173,
+      "step": 9540
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.4255299246700046e-05,
+      "loss": 1.9617,
+      "step": 9550
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.424301951532064e-05,
+      "loss": 1.9543,
+      "step": 9560
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.423072806629994e-05,
+      "loss": 1.9604,
+      "step": 9570
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.421842490557892e-05,
+      "loss": 1.952,
+      "step": 9580
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.420611003910419e-05,
+      "loss": 1.9601,
+      "step": 9590
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.4193783472828024e-05,
+      "loss": 1.9285,
+      "step": 9600
+    },
+    {
+      "epoch": 0.42,
+      "eval_loss": 1.922660231590271,
+      "eval_runtime": 11.9313,
+      "eval_samples_per_second": 343.299,
+      "eval_steps_per_second": 21.456,
+      "step": 9600
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.418144521270836e-05,
+      "loss": 1.9213,
+      "step": 9610
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.416909526470878e-05,
+      "loss": 1.9605,
+      "step": 9620
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.4156733634798535e-05,
+      "loss": 1.9544,
+      "step": 9630
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.414436032895248e-05,
+      "loss": 1.941,
+      "step": 9640
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.413197535315116e-05,
+      "loss": 1.9717,
+      "step": 9650
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.4119578713380726e-05,
+      "loss": 1.9784,
+      "step": 9660
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.410717041563298e-05,
+      "loss": 1.9701,
+      "step": 9670
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.409475046590537e-05,
+      "loss": 1.9148,
+      "step": 9680
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.408231887020095e-05,
+      "loss": 1.9512,
+      "step": 9690
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.406987563452844e-05,
+      "loss": 1.9492,
+      "step": 9700
+    },
+    {
+      "epoch": 0.42,
+      "eval_loss": 1.9206269979476929,
+      "eval_runtime": 11.8263,
+      "eval_samples_per_second": 346.347,
+      "eval_steps_per_second": 21.647,
+      "step": 9700
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.405742076490214e-05,
+      "loss": 1.9556,
+      "step": 9710
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.4044954267342005e-05,
+      "loss": 1.9188,
+      "step": 9720
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.403247614787361e-05,
+      "loss": 1.9446,
+      "step": 9730
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.4019986412528135e-05,
+      "loss": 1.913,
+      "step": 9740
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.400748506734237e-05,
+      "loss": 1.9685,
+      "step": 9750
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.3994972118358743e-05,
+      "loss": 1.8976,
+      "step": 9760
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.398244757162527e-05,
+      "loss": 1.9133,
+      "step": 9770
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.396991143319555e-05,
+      "loss": 1.9174,
+      "step": 9780
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.395736370912884e-05,
+      "loss": 1.9537,
+      "step": 9790
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.394480440548997e-05,
+      "loss": 1.9391,
+      "step": 9800
+    },
+    {
+      "epoch": 0.43,
+      "eval_loss": 1.9210110902786255,
+      "eval_runtime": 11.7383,
+      "eval_samples_per_second": 348.944,
+      "eval_steps_per_second": 21.809,
+      "step": 9800
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.393223352834933e-05,
+      "loss": 1.9293,
+      "step": 9810
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.391965108378298e-05,
+      "loss": 1.9753,
+      "step": 9820
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.3907057077872496e-05,
+      "loss": 1.921,
+      "step": 9830
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.38944515167051e-05,
+      "loss": 1.9402,
+      "step": 9840
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.388183440637356e-05,
+      "loss": 1.9688,
+      "step": 9850
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.3869205752976237e-05,
+      "loss": 1.9478,
+      "step": 9860
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.385656556261709e-05,
+      "loss": 1.9596,
+      "step": 9870
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.3843913841405637e-05,
+      "loss": 1.9397,
+      "step": 9880
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.3831250595456956e-05,
+      "loss": 1.9378,
+      "step": 9890
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.3818575830891716e-05,
+      "loss": 1.9586,
+      "step": 9900
+    },
+    {
+      "epoch": 0.43,
+      "eval_loss": 1.919926643371582,
+      "eval_runtime": 11.7669,
+      "eval_samples_per_second": 348.096,
+      "eval_steps_per_second": 21.756,
+      "step": 9900
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.380588955383616e-05,
+      "loss": 1.9818,
+      "step": 9910
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.379319177042208e-05,
+      "loss": 1.9407,
+      "step": 9920
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.378048248678682e-05,
+      "loss": 1.9797,
+      "step": 9930
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.3767761709073314e-05,
+      "loss": 1.9684,
+      "step": 9940
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.3755029443430024e-05,
+      "loss": 1.9438,
+      "step": 9950
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.374228569601098e-05,
+      "loss": 1.9506,
+      "step": 9960
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.372953047297574e-05,
+      "loss": 1.9582,
+      "step": 9970
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.371676378048944e-05,
+      "loss": 1.8836,
+      "step": 9980
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.370398562472273e-05,
+      "loss": 1.879,
+      "step": 9990
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.3691196011851835e-05,
+      "loss": 1.8977,
+      "step": 10000
+    },
+    {
+      "epoch": 0.44,
+      "eval_loss": 1.9170405864715576,
+      "eval_runtime": 12.0985,
+      "eval_samples_per_second": 338.555,
+      "eval_steps_per_second": 21.16,
+      "step": 10000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.367839494805847e-05,
+      "loss": 1.9891,
+      "step": 10010
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.3665582439529944e-05,
+      "loss": 1.9025,
+      "step": 10020
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.365275849245904e-05,
+      "loss": 1.9416,
+      "step": 10030
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.3639923113044107e-05,
+      "loss": 1.954,
+      "step": 10040
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.362707630748901e-05,
+      "loss": 1.9102,
+      "step": 10050
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.361421808200312e-05,
+      "loss": 1.9629,
+      "step": 10060
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.360134844280135e-05,
+      "loss": 1.9691,
+      "step": 10070
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.358846739610413e-05,
+      "loss": 1.9133,
+      "step": 10080
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.357557494813738e-05,
+      "loss": 1.9778,
+      "step": 10090
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.3562671105132566e-05,
+      "loss": 1.9306,
+      "step": 10100
+    },
+    {
+      "epoch": 0.44,
+      "eval_loss": 1.917136549949646,
+      "eval_runtime": 11.7385,
+      "eval_samples_per_second": 348.936,
+      "eval_steps_per_second": 21.808,
+      "step": 10100
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.3549755873326636e-05,
+      "loss": 1.9673,
+      "step": 10110
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.353682925896204e-05,
+      "loss": 1.9411,
+      "step": 10120
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.352389126828674e-05,
+      "loss": 1.9431,
+      "step": 10130
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.3510941907554205e-05,
+      "loss": 1.9393,
+      "step": 10140
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.349798118302338e-05,
+      "loss": 1.9358,
+      "step": 10150
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.348500910095873e-05,
+      "loss": 1.9342,
+      "step": 10160
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.3472025667630164e-05,
+      "loss": 1.9211,
+      "step": 10170
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.345903088931312e-05,
+      "loss": 1.9403,
+      "step": 10180
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.344602477228851e-05,
+      "loss": 1.9169,
+      "step": 10190
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.343300732284272e-05,
+      "loss": 1.9776,
+      "step": 10200
+    },
+    {
+      "epoch": 0.45,
+      "eval_loss": 1.9152053594589233,
+      "eval_runtime": 11.8351,
+      "eval_samples_per_second": 346.088,
+      "eval_steps_per_second": 21.631,
+      "step": 10200
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.34199785472676e-05,
+      "loss": 1.9555,
+      "step": 10210
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.340693845186051e-05,
+      "loss": 1.9321,
+      "step": 10220
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.339388704292425e-05,
+      "loss": 1.9974,
+      "step": 10230
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.33808243267671e-05,
+      "loss": 1.9198,
+      "step": 10240
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.33677503097028e-05,
+      "loss": 1.9572,
+      "step": 10250
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.335466499805057e-05,
+      "loss": 1.9275,
+      "step": 10260
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.334156839813506e-05,
+      "loss": 1.9221,
+      "step": 10270
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.33284605162864e-05,
+      "loss": 1.9517,
+      "step": 10280
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.3315341358840155e-05,
+      "loss": 1.9344,
+      "step": 10290
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.330221093213736e-05,
+      "loss": 1.9339,
+      "step": 10300
+    },
+    {
+      "epoch": 0.45,
+      "eval_loss": 1.9138355255126953,
+      "eval_runtime": 11.7912,
+      "eval_samples_per_second": 347.377,
+      "eval_steps_per_second": 21.711,
+      "step": 10300
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.3289069242524486e-05,
+      "loss": 1.942,
+      "step": 10310
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.327591629635345e-05,
+      "loss": 1.9285,
+      "step": 10320
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.32627520999816e-05,
+      "loss": 1.9394,
+      "step": 10330
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.3249576659771723e-05,
+      "loss": 1.9306,
+      "step": 10340
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.323638998209207e-05,
+      "loss": 1.9389,
+      "step": 10350
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.322319207331628e-05,
+      "loss": 1.9811,
+      "step": 10360
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.320998293982345e-05,
+      "loss": 1.9051,
+      "step": 10370
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.31967625879981e-05,
+      "loss": 1.9013,
+      "step": 10380
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.3183531024230145e-05,
+      "loss": 1.9472,
+      "step": 10390
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.3170288254914974e-05,
+      "loss": 1.9531,
+      "step": 10400
+    },
+    {
+      "epoch": 0.46,
+      "eval_loss": 1.9130414724349976,
+      "eval_runtime": 11.7623,
+      "eval_samples_per_second": 348.231,
+      "eval_steps_per_second": 21.764,
+      "step": 10400
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.315703428645332e-05,
+      "loss": 1.9145,
+      "step": 10410
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.31437691252514e-05,
+      "loss": 1.9428,
+      "step": 10420
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.3130492777720784e-05,
+      "loss": 1.9159,
+      "step": 10430
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.311720525027847e-05,
+      "loss": 1.9567,
+      "step": 10440
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.310390654934689e-05,
+      "loss": 1.9582,
+      "step": 10450
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.309059668135384e-05,
+      "loss": 1.8719,
+      "step": 10460
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.3077275652732506e-05,
+      "loss": 1.9593,
+      "step": 10470
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.3063943469921495e-05,
+      "loss": 1.9132,
+      "step": 10480
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.305060013936481e-05,
+      "loss": 1.9277,
+      "step": 10490
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.303724566751179e-05,
+      "loss": 1.9285,
+      "step": 10500
+    },
+    {
+      "epoch": 0.46,
+      "eval_loss": 1.912137746810913,
+      "eval_runtime": 11.8324,
+      "eval_samples_per_second": 346.169,
+      "eval_steps_per_second": 21.636,
+      "step": 10500
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.302388006081724e-05,
+      "loss": 1.9365,
+      "step": 10510
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.301050332574128e-05,
+      "loss": 1.9349,
+      "step": 10520
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.299711546874944e-05,
+      "loss": 1.9636,
+      "step": 10530
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.2983716496312595e-05,
+      "loss": 1.9193,
+      "step": 10540
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.297030641490705e-05,
+      "loss": 1.925,
+      "step": 10550
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.2956885231014426e-05,
+      "loss": 1.9379,
+      "step": 10560
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.294345295112173e-05,
+      "loss": 1.9314,
+      "step": 10570
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.2930009581721345e-05,
+      "loss": 1.9457,
+      "step": 10580
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.291655512931098e-05,
+      "loss": 1.9083,
+      "step": 10590
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.290308960039373e-05,
+      "loss": 1.9104,
+      "step": 10600
+    },
+    {
+      "epoch": 0.46,
+      "eval_loss": 1.9107956886291504,
+      "eval_runtime": 12.0695,
+      "eval_samples_per_second": 339.368,
+      "eval_steps_per_second": 21.21,
+      "step": 10600
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.288961300147804e-05,
+      "loss": 1.9385,
+      "step": 10610
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.287612533907769e-05,
+      "loss": 1.9446,
+      "step": 10620
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.286262661971183e-05,
+      "loss": 1.9284,
+      "step": 10630
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.284911684990494e-05,
+      "loss": 1.9348,
+      "step": 10640
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.283559603618683e-05,
+      "loss": 1.9606,
+      "step": 10650
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.2822064185092676e-05,
+      "loss": 1.883,
+      "step": 10660
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.280852130316297e-05,
+      "loss": 1.9772,
+      "step": 10670
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.279496739694355e-05,
+      "loss": 1.9849,
+      "step": 10680
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.2781402472985544e-05,
+      "loss": 1.916,
+      "step": 10690
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.276782653784546e-05,
+      "loss": 1.9241,
+      "step": 10700
+    },
+    {
+      "epoch": 0.47,
+      "eval_loss": 1.9091017246246338,
+      "eval_runtime": 11.9167,
+      "eval_samples_per_second": 343.72,
+      "eval_steps_per_second": 21.483,
+      "step": 10700
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.275423959808509e-05,
+      "loss": 1.9683,
+      "step": 10710
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.274064166027156e-05,
+      "loss": 1.9504,
+      "step": 10720
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.272703273097731e-05,
+      "loss": 1.9201,
+      "step": 10730
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.271341281678009e-05,
+      "loss": 1.9344,
+      "step": 10740
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.2699781924262966e-05,
+      "loss": 1.9063,
+      "step": 10750
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.2686140060014297e-05,
+      "loss": 1.8997,
+      "step": 10760
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.267248723062775e-05,
+      "loss": 1.8877,
+      "step": 10770
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.26588234427023e-05,
+      "loss": 1.9442,
+      "step": 10780
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.2645148702842224e-05,
+      "loss": 1.8964,
+      "step": 10790
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.2631463017657064e-05,
+      "loss": 1.8983,
+      "step": 10800
+    },
+    {
+      "epoch": 0.47,
+      "eval_loss": 1.9094958305358887,
+      "eval_runtime": 11.7143,
+      "eval_samples_per_second": 349.659,
+      "eval_steps_per_second": 21.854,
+      "step": 10800
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.261776639376169e-05,
+      "loss": 1.9515,
+      "step": 10810
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.260405883777622e-05,
+      "loss": 1.9609,
+      "step": 10820
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.259034035632607e-05,
+      "loss": 1.9301,
+      "step": 10830
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.2576610956041976e-05,
+      "loss": 1.946,
+      "step": 10840
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.2562870643559895e-05,
+      "loss": 1.9419,
+      "step": 10850
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.254911942552108e-05,
+      "loss": 1.9415,
+      "step": 10860
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.2535357308572064e-05,
+      "loss": 1.9428,
+      "step": 10870
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.252158429936464e-05,
+      "loss": 1.9382,
+      "step": 10880
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.250780040455586e-05,
+      "loss": 1.937,
+      "step": 10890
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.249400563080804e-05,
+      "loss": 1.9589,
+      "step": 10900
+    },
+    {
+      "epoch": 0.48,
+      "eval_loss": 1.9076216220855713,
+      "eval_runtime": 11.7896,
+      "eval_samples_per_second": 347.424,
+      "eval_steps_per_second": 21.714,
+      "step": 10900
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.2480199984788765e-05,
+      "loss": 1.919,
+      "step": 10910
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.246638347317086e-05,
+      "loss": 1.947,
+      "step": 10920
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.245255610263243e-05,
+      "loss": 1.965,
+      "step": 10930
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.243871787985678e-05,
+      "loss": 1.9391,
+      "step": 10940
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.242486881153251e-05,
+      "loss": 1.9027,
+      "step": 10950
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.241100890435342e-05,
+      "loss": 1.9201,
+      "step": 10960
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.239713816501859e-05,
+      "loss": 1.9278,
+      "step": 10970
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.23832566002323e-05,
+      "loss": 1.9501,
+      "step": 10980
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.2369364216704084e-05,
+      "loss": 1.9276,
+      "step": 10990
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.2355461021148695e-05,
+      "loss": 1.9232,
+      "step": 11000
+    },
+    {
+      "epoch": 0.48,
+      "eval_loss": 1.9063364267349243,
+      "eval_runtime": 11.7844,
+      "eval_samples_per_second": 347.577,
+      "eval_steps_per_second": 21.724,
+      "step": 11000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.2341547020286116e-05,
+      "loss": 1.946,
+      "step": 11010
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.232762222084155e-05,
+      "loss": 1.931,
+      "step": 11020
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.231368662954542e-05,
+      "loss": 1.944,
+      "step": 11030
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.2299740253133383e-05,
+      "loss": 1.9547,
+      "step": 11040
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.2285783098346265e-05,
+      "loss": 1.8769,
+      "step": 11050
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.227181517193015e-05,
+      "loss": 1.8999,
+      "step": 11060
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.22578364806363e-05,
+      "loss": 1.9489,
+      "step": 11070
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.2243847031221185e-05,
+      "loss": 1.9533,
+      "step": 11080
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.222984683044648e-05,
+      "loss": 1.9008,
+      "step": 11090
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.221583588507905e-05,
+      "loss": 1.9322,
+      "step": 11100
+    },
+    {
+      "epoch": 0.49,
+      "eval_loss": 1.905591607093811,
+      "eval_runtime": 11.9402,
+      "eval_samples_per_second": 343.044,
+      "eval_steps_per_second": 21.44,
+      "step": 11100
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.2201814201890964e-05,
+      "loss": 1.885,
+      "step": 11110
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.218778178765947e-05,
+      "loss": 1.9275,
+      "step": 11120
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.217373864916701e-05,
+      "loss": 1.9124,
+      "step": 11130
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.21596847932012e-05,
+      "loss": 1.9431,
+      "step": 11140
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.2145620226554844e-05,
+      "loss": 1.9124,
+      "step": 11150
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.213154495602593e-05,
+      "loss": 1.9301,
+      "step": 11160
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.211745898841759e-05,
+      "loss": 1.888,
+      "step": 11170
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.210336233053817e-05,
+      "loss": 1.9275,
+      "step": 11180
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.2089254989201156e-05,
+      "loss": 1.9612,
+      "step": 11190
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.207513697122519e-05,
+      "loss": 1.9247,
+      "step": 11200
+    },
+    {
+      "epoch": 0.49,
+      "eval_loss": 1.9045929908752441,
+      "eval_runtime": 11.6726,
+      "eval_samples_per_second": 350.906,
+      "eval_steps_per_second": 21.932,
+      "step": 11200
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.20610082834341e-05,
+      "loss": 1.9102,
+      "step": 11210
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.2046868932656855e-05,
+      "loss": 1.923,
+      "step": 11220
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.203271892572757e-05,
+      "loss": 1.8978,
+      "step": 11230
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.201855826948553e-05,
+      "loss": 1.8708,
+      "step": 11240
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.200438697077516e-05,
+      "loss": 1.9239,
+      "step": 11250
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.199020503644603e-05,
+      "loss": 1.9347,
+      "step": 11260
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.1976012473352834e-05,
+      "loss": 1.8907,
+      "step": 11270
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.196180928835543e-05,
+      "loss": 1.9238,
+      "step": 11280
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.1947595488318794e-05,
+      "loss": 1.925,
+      "step": 11290
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.1933371080113034e-05,
+      "loss": 1.8946,
+      "step": 11300
+    },
+    {
+      "epoch": 0.49,
+      "eval_loss": 1.9035365581512451,
+      "eval_runtime": 11.7334,
+      "eval_samples_per_second": 349.089,
+      "eval_steps_per_second": 21.818,
+      "step": 11300
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.191913607061339e-05,
+      "loss": 1.885,
+      "step": 11310
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.190489046670022e-05,
+      "loss": 1.9246,
+      "step": 11320
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.189063427525901e-05,
+      "loss": 1.9431,
+      "step": 11330
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.1876367503180356e-05,
+      "loss": 1.9154,
+      "step": 11340
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.186209015735998e-05,
+      "loss": 1.9025,
+      "step": 11350
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.184780224469869e-05,
+      "loss": 1.96,
+      "step": 11360
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.183350377210243e-05,
+      "loss": 1.9536,
+      "step": 11370
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.181919474648224e-05,
+      "loss": 1.9724,
+      "step": 11380
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.180487517475424e-05,
+      "loss": 1.8894,
+      "step": 11390
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.1790545063839675e-05,
+      "loss": 1.894,
+      "step": 11400
+    },
+    {
+      "epoch": 0.5,
+      "eval_loss": 1.9031916856765747,
+      "eval_runtime": 11.9751,
+      "eval_samples_per_second": 342.043,
+      "eval_steps_per_second": 21.378,
+      "step": 11400
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.177620442066487e-05,
+      "loss": 1.8651,
+      "step": 11410
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.176185325216124e-05,
+      "loss": 1.9263,
+      "step": 11420
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.1747491565265304e-05,
+      "loss": 1.9026,
+      "step": 11430
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.173311936691864e-05,
+      "loss": 1.9311,
+      "step": 11440
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.171873666406792e-05,
+      "loss": 1.9022,
+      "step": 11450
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.170434346366489e-05,
+      "loss": 1.9235,
+      "step": 11460
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.1689939772666376e-05,
+      "loss": 1.9091,
+      "step": 11470
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.1675525598034275e-05,
+      "loss": 1.9388,
+      "step": 11480
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.166110094673553e-05,
+      "loss": 1.9413,
+      "step": 11490
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.164666582574217e-05,
+      "loss": 1.9681,
+      "step": 11500
+    },
+    {
+      "epoch": 0.5,
+      "eval_loss": 1.9012072086334229,
+      "eval_runtime": 11.9665,
+      "eval_samples_per_second": 342.288,
+      "eval_steps_per_second": 21.393,
+      "step": 11500
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.163222024203129e-05,
+      "loss": 1.9637,
+      "step": 11510
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.161776420258502e-05,
+      "loss": 1.9221,
+      "step": 11520
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.1603297714390555e-05,
+      "loss": 1.9217,
+      "step": 11530
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.1588820784440144e-05,
+      "loss": 1.9271,
+      "step": 11540
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.157433341973107e-05,
+      "loss": 1.9115,
+      "step": 11550
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.155983562726568e-05,
+      "loss": 1.9195,
+      "step": 11560
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.154532741405133e-05,
+      "loss": 1.9535,
+      "step": 11570
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.153080878710046e-05,
+      "loss": 1.8912,
+      "step": 11580
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.151627975343049e-05,
+      "loss": 1.8922,
+      "step": 11590
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.1501740320063906e-05,
+      "loss": 1.8981,
+      "step": 11600
+    },
+    {
+      "epoch": 0.51,
+      "eval_loss": 1.9007313251495361,
+      "eval_runtime": 11.7722,
+      "eval_samples_per_second": 347.939,
+      "eval_steps_per_second": 21.746,
+      "step": 11600
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.148719049402821e-05,
+      "loss": 1.919,
+      "step": 11610
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.147263028235593e-05,
+      "loss": 1.8871,
+      "step": 11620
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.145805969208459e-05,
+      "loss": 1.9283,
+      "step": 11630
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.144347873025679e-05,
+      "loss": 1.9203,
+      "step": 11640
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.1428887403920064e-05,
+      "loss": 1.9441,
+      "step": 11650
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.141428572012703e-05,
+      "loss": 1.9235,
+      "step": 11660
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.1399673685935244e-05,
+      "loss": 1.8727,
+      "step": 11670
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.138505130840733e-05,
+      "loss": 1.9103,
+      "step": 11680
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.137041859461086e-05,
+      "loss": 1.9122,
+      "step": 11690
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.1355775551618435e-05,
+      "loss": 1.9197,
+      "step": 11700
+    },
+    {
+      "epoch": 0.51,
+      "eval_loss": 1.8997355699539185,
+      "eval_runtime": 11.7556,
+      "eval_samples_per_second": 348.43,
+      "eval_steps_per_second": 21.777,
+      "step": 11700
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.134112218650762e-05,
+      "loss": 1.9484,
+      "step": 11710
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.1326458506360994e-05,
+      "loss": 1.9234,
+      "step": 11720
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.131178451826612e-05,
+      "loss": 1.932,
+      "step": 11730
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.1297100229315536e-05,
+      "loss": 1.9191,
+      "step": 11740
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.128240564660673e-05,
+      "loss": 1.9287,
+      "step": 11750
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.1267700777242234e-05,
+      "loss": 1.8905,
+      "step": 11760
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.1252985628329485e-05,
+      "loss": 1.9224,
+      "step": 11770
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.123826020698092e-05,
+      "loss": 1.9458,
+      "step": 11780
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.122352452031394e-05,
+      "loss": 1.9364,
+      "step": 11790
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.1208778575450904e-05,
+      "loss": 1.8697,
+      "step": 11800
+    },
+    {
+      "epoch": 0.52,
+      "eval_loss": 1.8991934061050415,
+      "eval_runtime": 11.702,
+      "eval_samples_per_second": 350.026,
+      "eval_steps_per_second": 21.877,
+      "step": 11800
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.1194022379519134e-05,
+      "loss": 1.9021,
+      "step": 11810
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.1179255939650894e-05,
+      "loss": 1.9163,
+      "step": 11820
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.1164479262983405e-05,
+      "loss": 1.9285,
+      "step": 11830
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.1149692356658856e-05,
+      "loss": 1.9226,
+      "step": 11840
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.113489522782434e-05,
+      "loss": 1.8812,
+      "step": 11850
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.112008788363192e-05,
+      "loss": 1.9432,
+      "step": 11860
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.110527033123861e-05,
+      "loss": 1.8859,
+      "step": 11870
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.1090442577806306e-05,
+      "loss": 1.9347,
+      "step": 11880
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.107560463050188e-05,
+      "loss": 1.9167,
+      "step": 11890
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.106075649649714e-05,
+      "loss": 1.9169,
+      "step": 11900
+    },
+    {
+      "epoch": 0.52,
+      "eval_loss": 1.8978095054626465,
+      "eval_runtime": 11.7169,
+      "eval_samples_per_second": 349.58,
+      "eval_steps_per_second": 21.849,
+      "step": 11900
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.104589818296875e-05,
+      "loss": 1.9154,
+      "step": 11910
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.103102969709839e-05,
+      "loss": 1.9427,
+      "step": 11920
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.1016151046072576e-05,
+      "loss": 1.9245,
+      "step": 11930
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.100126223708276e-05,
+      "loss": 1.929,
+      "step": 11940
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.098636327732534e-05,
+      "loss": 1.912,
+      "step": 11950
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.097145417400157e-05,
+      "loss": 1.9093,
+      "step": 11960
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.0956534934317624e-05,
+      "loss": 1.9014,
+      "step": 11970
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.094160556548461e-05,
+      "loss": 1.9396,
+      "step": 11980
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.092666607471847e-05,
+      "loss": 1.9046,
+      "step": 11990
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.091171646924009e-05,
+      "loss": 1.9012,
+      "step": 12000
+    },
+    {
+      "epoch": 0.53,
+      "eval_loss": 1.897355556488037,
+      "eval_runtime": 12.4634,
+      "eval_samples_per_second": 328.642,
+      "eval_steps_per_second": 20.54,
+      "step": 12000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.089675675627522e-05,
+      "loss": 1.92,
+      "step": 12010
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.0881786943054486e-05,
+      "loss": 1.9191,
+      "step": 12020
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.086680703681343e-05,
+      "loss": 1.9002,
+      "step": 12030
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.085181704479244e-05,
+      "loss": 1.9084,
+      "step": 12040
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.0836816974236796e-05,
+      "loss": 1.9434,
+      "step": 12050
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.082180683239664e-05,
+      "loss": 1.9575,
+      "step": 12060
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.0806786626527e-05,
+      "loss": 1.9332,
+      "step": 12070
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.079175636388773e-05,
+      "loss": 1.8858,
+      "step": 12080
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.077671605174359e-05,
+      "loss": 1.8997,
+      "step": 12090
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.076166569736418e-05,
+      "loss": 1.9475,
+      "step": 12100
+    },
+    {
+      "epoch": 0.53,
+      "eval_loss": 1.8951764106750488,
+      "eval_runtime": 11.8716,
+      "eval_samples_per_second": 345.025,
+      "eval_steps_per_second": 21.564,
+      "step": 12100
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.074660530802393e-05,
+      "loss": 1.9078,
+      "step": 12110
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.073153489100216e-05,
+      "loss": 1.9249,
+      "step": 12120
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.0716454453583015e-05,
+      "loss": 1.8992,
+      "step": 12130
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.0701364003055475e-05,
+      "loss": 1.8958,
+      "step": 12140
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.0686263546713386e-05,
+      "loss": 1.9222,
+      "step": 12150
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.0671153091855406e-05,
+      "loss": 1.9098,
+      "step": 12160
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.0656032645785026e-05,
+      "loss": 1.8734,
+      "step": 12170
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.06409022158106e-05,
+      "loss": 1.9222,
+      "step": 12180
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.062576180924526e-05,
+      "loss": 1.9241,
+      "step": 12190
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.0610611433406996e-05,
+      "loss": 1.9312,
+      "step": 12200
+    },
+    {
+      "epoch": 0.53,
+      "eval_loss": 1.8951544761657715,
+      "eval_runtime": 11.9349,
+      "eval_samples_per_second": 343.194,
+      "eval_steps_per_second": 21.45,
+      "step": 12200
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.0595451095618595e-05,
+      "loss": 1.8868,
+      "step": 12210
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.058028080320767e-05,
+      "loss": 1.9428,
+      "step": 12220
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.056510056350665e-05,
+      "loss": 1.8836,
+      "step": 12230
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.0549910383852756e-05,
+      "loss": 1.9613,
+      "step": 12240
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.0534710271588026e-05,
+      "loss": 1.9218,
+      "step": 12250
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.051950023405928e-05,
+      "loss": 1.8936,
+      "step": 12260
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.0504280278618175e-05,
+      "loss": 1.9501,
+      "step": 12270
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.048905041262113e-05,
+      "loss": 1.9222,
+      "step": 12280
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.0473810643429346e-05,
+      "loss": 1.9357,
+      "step": 12290
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.0458560978408844e-05,
+      "loss": 1.9024,
+      "step": 12300
+    },
+    {
+      "epoch": 0.54,
+      "eval_loss": 1.894567608833313,
+      "eval_runtime": 11.8015,
+      "eval_samples_per_second": 347.074,
+      "eval_steps_per_second": 21.692,
+      "step": 12300
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.04433014249304e-05,
+      "loss": 1.901,
+      "step": 12310
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.042803199036958e-05,
+      "loss": 1.8953,
+      "step": 12320
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.041275268210672e-05,
+      "loss": 1.9358,
+      "step": 12330
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.039746350752696e-05,
+      "loss": 1.9632,
+      "step": 12340
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.0382164474020144e-05,
+      "loss": 1.8925,
+      "step": 12350
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.036685558898095e-05,
+      "loss": 1.8659,
+      "step": 12360
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.035153685980877e-05,
+      "loss": 1.9407,
+      "step": 12370
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.033620829390778e-05,
+      "loss": 1.886,
+      "step": 12380
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.0320869898686897e-05,
+      "loss": 1.8922,
+      "step": 12390
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.0305521681559795e-05,
+      "loss": 1.9246,
+      "step": 12400
+    },
+    {
+      "epoch": 0.54,
+      "eval_loss": 1.8931217193603516,
+      "eval_runtime": 11.8736,
+      "eval_samples_per_second": 344.967,
+      "eval_steps_per_second": 21.56,
+      "step": 12400
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.0290163649944895e-05,
+      "loss": 1.9126,
+      "step": 12410
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.0274795811265356e-05,
+      "loss": 1.8948,
+      "step": 12420
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.0259418172949096e-05,
+      "loss": 1.942,
+      "step": 12430
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.0244030742428746e-05,
+      "loss": 1.8984,
+      "step": 12440
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.022863352714168e-05,
+      "loss": 1.8838,
+      "step": 12450
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.0213226534529994e-05,
+      "loss": 1.8943,
+      "step": 12460
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.0197809772040526e-05,
+      "loss": 1.9087,
+      "step": 12470
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.0182383247124826e-05,
+      "loss": 1.9317,
+      "step": 12480
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.0166946967239155e-05,
+      "loss": 1.9581,
+      "step": 12490
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.0151500939844505e-05,
+      "loss": 1.9311,
+      "step": 12500
+    },
+    {
+      "epoch": 0.55,
+      "eval_loss": 1.8927887678146362,
+      "eval_runtime": 12.0202,
+      "eval_samples_per_second": 340.759,
+      "eval_steps_per_second": 21.297,
+      "step": 12500
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.013604517240657e-05,
+      "loss": 1.9014,
+      "step": 12510
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.012057967239576e-05,
+      "loss": 1.8831,
+      "step": 12520
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.010510444728717e-05,
+      "loss": 1.9165,
+      "step": 12530
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.008961950456062e-05,
+      "loss": 1.8805,
+      "step": 12540
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.0074124851700615e-05,
+      "loss": 1.8601,
+      "step": 12550
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.005862049619634e-05,
+      "loss": 1.9155,
+      "step": 12560
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.00431064455417e-05,
+      "loss": 1.923,
+      "step": 12570
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.002758270723525e-05,
+      "loss": 1.9147,
+      "step": 12580
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.0012049288780266e-05,
+      "loss": 1.9428,
+      "step": 12590
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.9996506197684674e-05,
+      "loss": 1.9652,
+      "step": 12600
+    },
+    {
+      "epoch": 0.55,
+      "eval_loss": 1.8919925689697266,
+      "eval_runtime": 11.9961,
+      "eval_samples_per_second": 341.443,
+      "eval_steps_per_second": 21.34,
+      "step": 12600
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.9980953441461076e-05,
+      "loss": 1.9505,
+      "step": 12610
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.9965391027626776e-05,
+      "loss": 1.8803,
+      "step": 12620
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.99498189637037e-05,
+      "loss": 1.9102,
+      "step": 12630
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.993423725721849e-05,
+      "loss": 1.9154,
+      "step": 12640
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.99186459157024e-05,
+      "loss": 1.9074,
+      "step": 12650
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.9903044946691354e-05,
+      "loss": 1.9372,
+      "step": 12660
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.988743435772596e-05,
+      "loss": 1.9326,
+      "step": 12670
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.9871814156351444e-05,
+      "loss": 1.8712,
+      "step": 12680
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.9856184350117696e-05,
+      "loss": 1.9461,
+      "step": 12690
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.9840544946579226e-05,
+      "loss": 1.9164,
+      "step": 12700
+    },
+    {
+      "epoch": 0.56,
+      "eval_loss": 1.8904156684875488,
+      "eval_runtime": 11.9521,
+      "eval_samples_per_second": 342.701,
+      "eval_steps_per_second": 21.419,
+      "step": 12700
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.98248959532952e-05,
+      "loss": 1.8621,
+      "step": 12710
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.980923737782941e-05,
+      "loss": 1.8796,
+      "step": 12720
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.979356922775029e-05,
+      "loss": 1.9101,
+      "step": 12730
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.9777891510630904e-05,
+      "loss": 1.9218,
+      "step": 12740
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.976220423404894e-05,
+      "loss": 1.904,
+      "step": 12750
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.9746507405586664e-05,
+      "loss": 1.947,
+      "step": 12760
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.973080103283103e-05,
+      "loss": 1.9095,
+      "step": 12770
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.971508512337355e-05,
+      "loss": 1.8795,
+      "step": 12780
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.969935968481037e-05,
+      "loss": 1.9002,
+      "step": 12790
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.9683624724742246e-05,
+      "loss": 1.9281,
+      "step": 12800
+    },
+    {
+      "epoch": 0.56,
+      "eval_loss": 1.8897275924682617,
+      "eval_runtime": 11.5813,
+      "eval_samples_per_second": 353.674,
+      "eval_steps_per_second": 22.105,
+      "step": 12800
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.96678802507745e-05,
+      "loss": 1.875,
+      "step": 12810
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.965212627051712e-05,
+      "loss": 1.8754,
+      "step": 12820
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.9636362791584606e-05,
+      "loss": 1.8657,
+      "step": 12830
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.9620589821596115e-05,
+      "loss": 1.922,
+      "step": 12840
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.960480736817537e-05,
+      "loss": 1.9201,
+      "step": 12850
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.958901543895066e-05,
+      "loss": 1.8708,
+      "step": 12860
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.957321404155488e-05,
+      "loss": 1.9147,
+      "step": 12870
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.955740318362548e-05,
+      "loss": 1.9418,
+      "step": 12880
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.954158287280452e-05,
+      "loss": 1.9568,
+      "step": 12890
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.9525753116738566e-05,
+      "loss": 1.9146,
+      "step": 12900
+    },
+    {
+      "epoch": 0.56,
+      "eval_loss": 1.89072847366333,
+      "eval_runtime": 11.6591,
+      "eval_samples_per_second": 351.315,
+      "eval_steps_per_second": 21.957,
+      "step": 12900
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.950991392307881e-05,
+      "loss": 1.9184,
+      "step": 12910
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.949406529948097e-05,
+      "loss": 1.9761,
+      "step": 12920
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.947820725360534e-05,
+      "loss": 1.9419,
+      "step": 12930
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.946233979311676e-05,
+      "loss": 1.893,
+      "step": 12940
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.9446462925684616e-05,
+      "loss": 1.9044,
+      "step": 12950
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.943057665898285e-05,
+      "loss": 1.9326,
+      "step": 12960
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.941468100068994e-05,
+      "loss": 1.9218,
+      "step": 12970
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.9398775958488914e-05,
+      "loss": 1.9269,
+      "step": 12980
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.9382861540067324e-05,
+      "loss": 1.9444,
+      "step": 12990
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.9366937753117266e-05,
+      "loss": 1.8917,
+      "step": 13000
+    },
+    {
+      "epoch": 0.57,
+      "eval_loss": 1.888537883758545,
+      "eval_runtime": 11.8673,
+      "eval_samples_per_second": 345.151,
+      "eval_steps_per_second": 21.572,
+      "step": 13000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.9351004605335335e-05,
+      "loss": 1.8794,
+      "step": 13010
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.9335062104422693e-05,
+      "loss": 1.977,
+      "step": 13020
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.9319110258085e-05,
+      "loss": 1.918,
+      "step": 13030
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.930314907403243e-05,
+      "loss": 1.9226,
+      "step": 13040
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.928717855997966e-05,
+      "loss": 1.8847,
+      "step": 13050
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.927119872364591e-05,
+      "loss": 1.9315,
+      "step": 13060
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.925520957275489e-05,
+      "loss": 1.9221,
+      "step": 13070
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.9239211115034804e-05,
+      "loss": 1.9149,
+      "step": 13080
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.922320335821835e-05,
+      "loss": 1.8801,
+      "step": 13090
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.920718631004275e-05,
+      "loss": 1.8859,
+      "step": 13100
+    },
+    {
+      "epoch": 0.57,
+      "eval_loss": 1.8876938819885254,
+      "eval_runtime": 11.8093,
+      "eval_samples_per_second": 346.846,
+      "eval_steps_per_second": 21.678,
+      "step": 13100
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.9191159978249686e-05,
+      "loss": 1.9131,
+      "step": 13110
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.917512437058534e-05,
+      "loss": 1.9271,
+      "step": 13120
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.915907949480037e-05,
+      "loss": 1.9257,
+      "step": 13130
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.914302535864993e-05,
+      "loss": 1.9172,
+      "step": 13140
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.9126961969893625e-05,
+      "loss": 1.9085,
+      "step": 13150
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.911088933629557e-05,
+      "loss": 1.916,
+      "step": 13160
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.90948074656243e-05,
+      "loss": 1.9169,
+      "step": 13170
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.907871636565285e-05,
+      "loss": 1.9196,
+      "step": 13180
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.906261604415872e-05,
+      "loss": 1.919,
+      "step": 13190
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.904650650892384e-05,
+      "loss": 1.9009,
+      "step": 13200
+    },
+    {
+      "epoch": 0.58,
+      "eval_loss": 1.8867980241775513,
+      "eval_runtime": 11.7952,
+      "eval_samples_per_second": 347.26,
+      "eval_steps_per_second": 21.704,
+      "step": 13200
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.90303877677346e-05,
+      "loss": 1.8787,
+      "step": 13210
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.901425982838185e-05,
+      "loss": 1.9349,
+      "step": 13220
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.8998122698660884e-05,
+      "loss": 1.955,
+      "step": 13230
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.8981976386371444e-05,
+      "loss": 1.9376,
+      "step": 13240
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.896582089931768e-05,
+      "loss": 1.9519,
+      "step": 13250
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.894965624530821e-05,
+      "loss": 1.9243,
+      "step": 13260
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.8933482432156094e-05,
+      "loss": 1.9578,
+      "step": 13270
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.891729946767876e-05,
+      "loss": 1.9111,
+      "step": 13280
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.8901107359698115e-05,
+      "loss": 1.9282,
+      "step": 13290
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.8884906116040464e-05,
+      "loss": 1.8958,
+      "step": 13300
+    },
+    {
+      "epoch": 0.58,
+      "eval_loss": 1.8861165046691895,
+      "eval_runtime": 11.9112,
+      "eval_samples_per_second": 343.879,
+      "eval_steps_per_second": 21.492,
+      "step": 13300
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.886869574453653e-05,
+      "loss": 1.8374,
+      "step": 13310
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.8852476253021435e-05,
+      "loss": 1.9007,
+      "step": 13320
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.883624764933474e-05,
+      "loss": 1.9312,
+      "step": 13330
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.882000994132039e-05,
+      "loss": 1.9262,
+      "step": 13340
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.8803763136826715e-05,
+      "loss": 1.888,
+      "step": 13350
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.878750724370647e-05,
+      "loss": 1.9391,
+      "step": 13360
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.87712422698168e-05,
+      "loss": 1.8841,
+      "step": 13370
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.875496822301922e-05,
+      "loss": 1.8599,
+      "step": 13380
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.873868511117964e-05,
+      "loss": 1.8711,
+      "step": 13390
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.8722392942168365e-05,
+      "loss": 1.9013,
+      "step": 13400
+    },
+    {
+      "epoch": 0.59,
+      "eval_loss": 1.8849880695343018,
+      "eval_runtime": 11.6646,
+      "eval_samples_per_second": 351.149,
+      "eval_steps_per_second": 21.947,
+      "step": 13400
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.870609172386006e-05,
+      "loss": 1.8912,
+      "step": 13410
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.868978146413376e-05,
+      "loss": 1.8703,
+      "step": 13420
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.867346217087289e-05,
+      "loss": 1.92,
+      "step": 13430
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.865713385196522e-05,
+      "loss": 1.8996,
+      "step": 13440
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.8640796515302915e-05,
+      "loss": 1.9156,
+      "step": 13450
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.862445016878245e-05,
+      "loss": 1.927,
+      "step": 13460
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.8608094820304704e-05,
+      "loss": 1.9362,
+      "step": 13470
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.859173047777488e-05,
+      "loss": 1.8748,
+      "step": 13480
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.857535714910252e-05,
+      "loss": 1.9397,
+      "step": 13490
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.855897484220153e-05,
+      "loss": 1.8975,
+      "step": 13500
+    },
+    {
+      "epoch": 0.59,
+      "eval_loss": 1.8842600584030151,
+      "eval_runtime": 11.9977,
+      "eval_samples_per_second": 341.399,
+      "eval_steps_per_second": 21.337,
+      "step": 13500
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.854258356499016e-05,
+      "loss": 1.8826,
+      "step": 13510
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.8526183325390956e-05,
+      "loss": 1.8971,
+      "step": 13520
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.850977413133085e-05,
+      "loss": 1.9155,
+      "step": 13530
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.849335599074106e-05,
+      "loss": 1.9324,
+      "step": 13540
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.847692891155716e-05,
+      "loss": 1.8948,
+      "step": 13550
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.8460492901719006e-05,
+      "loss": 1.9081,
+      "step": 13560
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.84440479691708e-05,
+      "loss": 1.9164,
+      "step": 13570
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.842759412186107e-05,
+      "loss": 1.9046,
+      "step": 13580
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.8411131367742606e-05,
+      "loss": 1.8918,
+      "step": 13590
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.839465971477255e-05,
+      "loss": 1.9471,
+      "step": 13600
+    },
+    {
+      "epoch": 0.6,
+      "eval_loss": 1.8836712837219238,
+      "eval_runtime": 11.9691,
+      "eval_samples_per_second": 342.213,
+      "eval_steps_per_second": 21.388,
+      "step": 13600
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.8378179170912295e-05,
+      "loss": 1.9032,
+      "step": 13610
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.83616897441276e-05,
+      "loss": 1.9068,
+      "step": 13620
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.8345191442388444e-05,
+      "loss": 1.9466,
+      "step": 13630
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.832868427366915e-05,
+      "loss": 1.9245,
+      "step": 13640
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.83121682459483e-05,
+      "loss": 1.9372,
+      "step": 13650
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.829564336720877e-05,
+      "loss": 1.9026,
+      "step": 13660
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.827910964543769e-05,
+      "loss": 1.8805,
+      "step": 13670
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.82625670886265e-05,
+      "loss": 1.9085,
+      "step": 13680
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.82460157047709e-05,
+      "loss": 1.8671,
+      "step": 13690
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.822945550187083e-05,
+      "loss": 1.9411,
+      "step": 13700
+    },
+    {
+      "epoch": 0.6,
+      "eval_loss": 1.8830509185791016,
+      "eval_runtime": 11.7716,
+      "eval_samples_per_second": 347.957,
+      "eval_steps_per_second": 21.747,
+      "step": 13700
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.8212886487930526e-05,
+      "loss": 1.9245,
+      "step": 13710
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.819630867095845e-05,
+      "loss": 1.9302,
+      "step": 13720
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.817972205896738e-05,
+      "loss": 1.8732,
+      "step": 13730
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.816312665997426e-05,
+      "loss": 1.888,
+      "step": 13740
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.814652248200035e-05,
+      "loss": 1.8778,
+      "step": 13750
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.8129909533071105e-05,
+      "loss": 1.8716,
+      "step": 13760
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.811328782121626e-05,
+      "loss": 1.8734,
+      "step": 13770
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.809665735446975e-05,
+      "loss": 1.9905,
+      "step": 13780
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.8080018140869775e-05,
+      "loss": 1.9574,
+      "step": 13790
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.806337018845875e-05,
+      "loss": 1.8946,
+      "step": 13800
+    },
+    {
+      "epoch": 0.6,
+      "eval_loss": 1.8828606605529785,
+      "eval_runtime": 11.6574,
+      "eval_samples_per_second": 351.364,
+      "eval_steps_per_second": 21.96,
+      "step": 13800
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.804671350528329e-05,
+      "loss": 1.911,
+      "step": 13810
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.8030048099394265e-05,
+      "loss": 1.9134,
+      "step": 13820
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.801337397884675e-05,
+      "loss": 1.9293,
+      "step": 13830
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.799669115170001e-05,
+      "loss": 1.9327,
+      "step": 13840
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.797999962601755e-05,
+      "loss": 1.8865,
+      "step": 13850
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.796329940986706e-05,
+      "loss": 1.9236,
+      "step": 13860
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.794659051132044e-05,
+      "loss": 1.8992,
+      "step": 13870
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.7929872938453796e-05,
+      "loss": 1.913,
+      "step": 13880
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.791314669934739e-05,
+      "loss": 1.8726,
+      "step": 13890
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.78964118020857e-05,
+      "loss": 1.8791,
+      "step": 13900
+    },
+    {
+      "epoch": 0.61,
+      "eval_loss": 1.8819687366485596,
+      "eval_runtime": 11.7317,
+      "eval_samples_per_second": 349.139,
+      "eval_steps_per_second": 21.821,
+      "step": 13900
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.7879668254757404e-05,
+      "loss": 1.8488,
+      "step": 13910
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.786291606545533e-05,
+      "loss": 1.9093,
+      "step": 13920
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.784615524227648e-05,
+      "loss": 1.9286,
+      "step": 13930
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.782938579332207e-05,
+      "loss": 1.9215,
+      "step": 13940
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.7812607726697446e-05,
+      "loss": 1.9186,
+      "step": 13950
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.779582105051214e-05,
+      "loss": 1.8571,
+      "step": 13960
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.777902577287983e-05,
+      "loss": 1.9116,
+      "step": 13970
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.7762221901918364e-05,
+      "loss": 1.8792,
+      "step": 13980
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.7745409445749737e-05,
+      "loss": 1.8778,
+      "step": 13990
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.7728588412500095e-05,
+      "loss": 1.9064,
+      "step": 14000
+    },
+    {
+      "epoch": 0.61,
+      "eval_loss": 1.88134765625,
+      "eval_runtime": 12.1415,
+      "eval_samples_per_second": 337.357,
+      "eval_steps_per_second": 21.085,
+      "step": 14000
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.771175881029973e-05,
+      "loss": 1.9157,
+      "step": 14010
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.769492064728309e-05,
+      "loss": 1.8532,
+      "step": 14020
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.7678073931588716e-05,
+      "loss": 1.909,
+      "step": 14030
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.766121867135935e-05,
+      "loss": 1.9097,
+      "step": 14040
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.7644354874741795e-05,
+      "loss": 1.9228,
+      "step": 14050
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.762748254988704e-05,
+      "loss": 1.8766,
+      "step": 14060
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.7610601704950154e-05,
+      "loss": 1.8731,
+      "step": 14070
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.7593712348090335e-05,
+      "loss": 1.8987,
+      "step": 14080
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.75768144874709e-05,
+      "loss": 1.8735,
+      "step": 14090
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.755990813125929e-05,
+      "loss": 1.9096,
+      "step": 14100
+    },
+    {
+      "epoch": 0.62,
+      "eval_loss": 1.8808305263519287,
+      "eval_runtime": 11.6253,
+      "eval_samples_per_second": 352.334,
+      "eval_steps_per_second": 22.021,
+      "step": 14100
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.754299328762703e-05,
+      "loss": 1.8673,
+      "step": 14110
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.7526069964749745e-05,
+      "loss": 1.9113,
+      "step": 14120
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.750913817080718e-05,
+      "loss": 1.906,
+      "step": 14130
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.749219791398315e-05,
+      "loss": 1.8916,
+      "step": 14140
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.747524920246558e-05,
+      "loss": 1.9327,
+      "step": 14150
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.745829204444648e-05,
+      "loss": 1.9021,
+      "step": 14160
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.744132644812192e-05,
+      "loss": 1.859,
+      "step": 14170
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.742435242169208e-05,
+      "loss": 1.8952,
+      "step": 14180
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.74073699733612e-05,
+      "loss": 1.9064,
+      "step": 14190
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.73903791113376e-05,
+      "loss": 1.8784,
+      "step": 14200
+    },
+    {
+      "epoch": 0.62,
+      "eval_loss": 1.878967523574829,
+      "eval_runtime": 11.7525,
+      "eval_samples_per_second": 348.52,
+      "eval_steps_per_second": 21.783,
+      "step": 14200
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.737337984383363e-05,
+      "loss": 1.8813,
+      "step": 14210
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.735637217906574e-05,
+      "loss": 1.8586,
+      "step": 14220
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.733935612525444e-05,
+      "loss": 1.9097,
+      "step": 14230
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.732233169062428e-05,
+      "loss": 1.9324,
+      "step": 14240
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.730529888340386e-05,
+      "loss": 1.9022,
+      "step": 14250
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.7288257711825836e-05,
+      "loss": 1.8895,
+      "step": 14260
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.7271208184126895e-05,
+      "loss": 1.9084,
+      "step": 14270
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.725415030854777e-05,
+      "loss": 1.8941,
+      "step": 14280
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.7237084093333244e-05,
+      "loss": 1.8961,
+      "step": 14290
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.72200095467321e-05,
+      "loss": 1.881,
+      "step": 14300
+    },
+    {
+      "epoch": 0.63,
+      "eval_loss": 1.8786323070526123,
+      "eval_runtime": 11.735,
+      "eval_samples_per_second": 349.042,
+      "eval_steps_per_second": 21.815,
+      "step": 14300
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.720292667699717e-05,
+      "loss": 1.9209,
+      "step": 14310
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.7185835492385294e-05,
+      "loss": 1.8944,
+      "step": 14320
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.716873600115736e-05,
+      "loss": 1.9029,
+      "step": 14330
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.7151628211578226e-05,
+      "loss": 1.9425,
+      "step": 14340
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.71345121319168e-05,
+      "loss": 1.9491,
+      "step": 14350
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.711738777044598e-05,
+      "loss": 1.9051,
+      "step": 14360
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.710025513544266e-05,
+      "loss": 1.9132,
+      "step": 14370
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.708311423518776e-05,
+      "loss": 1.9294,
+      "step": 14380
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.706596507796616e-05,
+      "loss": 1.8892,
+      "step": 14390
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.7048807672066754e-05,
+      "loss": 1.9103,
+      "step": 14400
+    },
+    {
+      "epoch": 0.63,
+      "eval_loss": 1.8775782585144043,
+      "eval_runtime": 11.7984,
+      "eval_samples_per_second": 347.167,
+      "eval_steps_per_second": 21.698,
+      "step": 14400
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.7031642025782416e-05,
+      "loss": 1.8831,
+      "step": 14410
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.701446814741001e-05,
+      "loss": 1.922,
+      "step": 14420
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.699728604525037e-05,
+      "loss": 1.9376,
+      "step": 14430
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.69800957276083e-05,
+      "loss": 1.9187,
+      "step": 14440
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.696289720279259e-05,
+      "loss": 1.8812,
+      "step": 14450
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.694569047911599e-05,
+      "loss": 1.9076,
+      "step": 14460
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.69284755648952e-05,
+      "loss": 1.9382,
+      "step": 14470
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.69112524684509e-05,
+      "loss": 1.8788,
+      "step": 14480
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.689402119810773e-05,
+      "loss": 1.9089,
+      "step": 14490
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.687678176219424e-05,
+      "loss": 1.913,
+      "step": 14500
+    },
+    {
+      "epoch": 0.63,
+      "eval_loss": 1.8763041496276855,
+      "eval_runtime": 12.2143,
+      "eval_samples_per_second": 335.344,
+      "eval_steps_per_second": 20.959,
+      "step": 14500
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.6859534169042976e-05,
+      "loss": 1.878,
+      "step": 14510
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.6842278426990397e-05,
+      "loss": 1.9019,
+      "step": 14520
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.68250145443769e-05,
+      "loss": 1.8764,
+      "step": 14530
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.6807742529546815e-05,
+      "loss": 1.9188,
+      "step": 14540
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.679046239084845e-05,
+      "loss": 1.9146,
+      "step": 14550
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.677317413663397e-05,
+      "loss": 1.8906,
+      "step": 14560
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.675587777525949e-05,
+      "loss": 1.875,
+      "step": 14570
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.6738573315085075e-05,
+      "loss": 1.9221,
+      "step": 14580
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.672126076447466e-05,
+      "loss": 1.928,
+      "step": 14590
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.670394013179611e-05,
+      "loss": 1.9176,
+      "step": 14600
+    },
+    {
+      "epoch": 0.64,
+      "eval_loss": 1.876225471496582,
+      "eval_runtime": 11.7678,
+      "eval_samples_per_second": 348.067,
+      "eval_steps_per_second": 21.754,
+      "step": 14600
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.66866114254212e-05,
+      "loss": 1.8598,
+      "step": 14610
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.666927465372559e-05,
+      "loss": 1.9218,
+      "step": 14620
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.665192982508884e-05,
+      "loss": 1.8575,
+      "step": 14630
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.6634576947894456e-05,
+      "loss": 1.9287,
+      "step": 14640
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.6617216030529746e-05,
+      "loss": 1.9112,
+      "step": 14650
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.659984708138597e-05,
+      "loss": 1.8923,
+      "step": 14660
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.658247010885826e-05,
+      "loss": 1.8565,
+      "step": 14670
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.6565085121345606e-05,
+      "loss": 1.8667,
+      "step": 14680
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.654769212725088e-05,
+      "loss": 1.8877,
+      "step": 14690
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.6530291134980825e-05,
+      "loss": 1.9241,
+      "step": 14700
+    },
+    {
+      "epoch": 0.64,
+      "eval_loss": 1.8761088848114014,
+      "eval_runtime": 12.0138,
+      "eval_samples_per_second": 340.942,
+      "eval_steps_per_second": 21.309,
+      "step": 14700
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.651288215294606e-05,
+      "loss": 1.8847,
+      "step": 14710
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.649546518956105e-05,
+      "loss": 1.8731,
+      "step": 14720
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.647804025324413e-05,
+      "loss": 1.8855,
+      "step": 14730
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.6460607352417476e-05,
+      "loss": 1.89,
+      "step": 14740
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.644316649550712e-05,
+      "loss": 1.8096,
+      "step": 14750
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.642571769094296e-05,
+      "loss": 1.9138,
+      "step": 14760
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.6408260947158684e-05,
+      "loss": 1.9277,
+      "step": 14770
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.6390796272591884e-05,
+      "loss": 1.8945,
+      "step": 14780
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.637332367568392e-05,
+      "loss": 1.8884,
+      "step": 14790
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.635584316488003e-05,
+      "loss": 1.9179,
+      "step": 14800
+    },
+    {
+      "epoch": 0.65,
+      "eval_loss": 1.8754827976226807,
+      "eval_runtime": 12.2675,
+      "eval_samples_per_second": 333.89,
+      "eval_steps_per_second": 20.868,
+      "step": 14800
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.6338354748629244e-05,
+      "loss": 1.9361,
+      "step": 14810
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.6320858435384446e-05,
+      "loss": 1.9061,
+      "step": 14820
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.630335423360232e-05,
+      "loss": 1.9,
+      "step": 14830
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.628584215174333e-05,
+      "loss": 1.8648,
+      "step": 14840
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.6268322198271804e-05,
+      "loss": 1.8879,
+      "step": 14850
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.625079438165585e-05,
+      "loss": 1.8491,
+      "step": 14860
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.6233258710367375e-05,
+      "loss": 1.867,
+      "step": 14870
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.621571519288209e-05,
+      "loss": 1.8921,
+      "step": 14880
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.619816383767949e-05,
+      "loss": 1.9075,
+      "step": 14890
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.6180604653242855e-05,
+      "loss": 1.9128,
+      "step": 14900
+    },
+    {
+      "epoch": 0.65,
+      "eval_loss": 1.874982476234436,
+      "eval_runtime": 11.9034,
+      "eval_samples_per_second": 344.103,
+      "eval_steps_per_second": 21.506,
+      "step": 14900
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.6163037648059256e-05,
+      "loss": 1.9357,
+      "step": 14910
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.614546283061955e-05,
+      "loss": 1.8716,
+      "step": 14920
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.612788020941837e-05,
+      "loss": 1.886,
+      "step": 14930
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.611028979295411e-05,
+      "loss": 1.9695,
+      "step": 14940
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.6092691589728924e-05,
+      "loss": 1.8555,
+      "step": 14950
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.607508560824876e-05,
+      "loss": 1.9099,
+      "step": 14960
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.6057471857023306e-05,
+      "loss": 1.8963,
+      "step": 14970
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.603985034456599e-05,
+      "loss": 1.8679,
+      "step": 14980
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.602222107939403e-05,
+      "loss": 1.9101,
+      "step": 14990
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.6004584070028354e-05,
+      "loss": 1.9419,
+      "step": 15000
+    },
+    {
+      "epoch": 0.66,
+      "eval_loss": 1.874194622039795,
+      "eval_runtime": 11.8292,
+      "eval_samples_per_second": 346.263,
+      "eval_steps_per_second": 21.641,
+      "step": 15000
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.598693932499366e-05,
+      "loss": 1.9005,
+      "step": 15010
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.596928685281836e-05,
+      "loss": 1.8898,
+      "step": 15020
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.5951626662034636e-05,
+      "loss": 1.8748,
+      "step": 15030
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.5933958761178355e-05,
+      "loss": 1.9403,
+      "step": 15040
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.5916283158789146e-05,
+      "loss": 1.8801,
+      "step": 15050
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.5898599863410355e-05,
+      "loss": 1.8711,
+      "step": 15060
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.5880908883589044e-05,
+      "loss": 1.8567,
+      "step": 15070
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.5863210227875965e-05,
+      "loss": 1.9125,
+      "step": 15080
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.584550390482562e-05,
+      "loss": 1.9104,
+      "step": 15090
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.582778992299618e-05,
+      "loss": 1.8765,
+      "step": 15100
+    },
+    {
+      "epoch": 0.66,
+      "eval_loss": 1.8737382888793945,
+      "eval_runtime": 11.8232,
+      "eval_samples_per_second": 346.437,
+      "eval_steps_per_second": 21.652,
+      "step": 15100
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.5810068290949566e-05,
+      "loss": 1.9292,
+      "step": 15110
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.5792339017251336e-05,
+      "loss": 1.9138,
+      "step": 15120
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.577460211047078e-05,
+      "loss": 1.8812,
+      "step": 15130
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.5756857579180887e-05,
+      "loss": 1.875,
+      "step": 15140
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.573910543195829e-05,
+      "loss": 1.8958,
+      "step": 15150
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.572134567738334e-05,
+      "loss": 1.9032,
+      "step": 15160
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.5703578324040044e-05,
+      "loss": 1.9176,
+      "step": 15170
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.56858033805161e-05,
+      "loss": 1.8699,
+      "step": 15180
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.5668020855402844e-05,
+      "loss": 1.8832,
+      "step": 15190
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.565023075729532e-05,
+      "loss": 1.9259,
+      "step": 15200
+    },
+    {
+      "epoch": 0.67,
+      "eval_loss": 1.872511625289917,
+      "eval_runtime": 11.6329,
+      "eval_samples_per_second": 352.105,
+      "eval_steps_per_second": 22.007,
+      "step": 15200
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.56324330947922e-05,
+      "loss": 1.8409,
+      "step": 15210
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.5614627876495825e-05,
+      "loss": 1.9122,
+      "step": 15220
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.559681511101217e-05,
+      "loss": 1.903,
+      "step": 15230
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.5578994806950876e-05,
+      "loss": 1.8904,
+      "step": 15240
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.556116697292524e-05,
+      "loss": 1.9553,
+      "step": 15250
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.554333161755216e-05,
+      "loss": 1.8666,
+      "step": 15260
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.552548874945221e-05,
+      "loss": 1.8376,
+      "step": 15270
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.550763837724957e-05,
+      "loss": 1.8787,
+      "step": 15280
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.5489780509572044e-05,
+      "loss": 1.9157,
+      "step": 15290
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.5471915155051084e-05,
+      "loss": 1.9041,
+      "step": 15300
+    },
+    {
+      "epoch": 0.67,
+      "eval_loss": 1.8725992441177368,
+      "eval_runtime": 11.7485,
+      "eval_samples_per_second": 348.639,
+      "eval_steps_per_second": 21.79,
+      "step": 15300
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.545404232232174e-05,
+      "loss": 1.9036,
+      "step": 15310
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.5436162020022686e-05,
+      "loss": 1.8459,
+      "step": 15320
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.541827425679618e-05,
+      "loss": 1.8782,
+      "step": 15330
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.540037904128814e-05,
+      "loss": 1.921,
+      "step": 15340
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.538247638214804e-05,
+      "loss": 1.913,
+      "step": 15350
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.536456628802895e-05,
+      "loss": 1.8882,
+      "step": 15360
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.5346648767587574e-05,
+      "loss": 1.9248,
+      "step": 15370
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.532872382948418e-05,
+      "loss": 1.9775,
+      "step": 15380
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.5310791482382616e-05,
+      "loss": 1.9272,
+      "step": 15390
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.52928517349503e-05,
+      "loss": 1.8919,
+      "step": 15400
+    },
+    {
+      "epoch": 0.67,
+      "eval_loss": 1.8721492290496826,
+      "eval_runtime": 11.7624,
+      "eval_samples_per_second": 348.228,
+      "eval_steps_per_second": 21.764,
+      "step": 15400
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.527490459585828e-05,
+      "loss": 1.902,
+      "step": 15410
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.525695007378112e-05,
+      "loss": 1.8867,
+      "step": 15420
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.523898817739697e-05,
+      "loss": 1.8574,
+      "step": 15430
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.522101891538755e-05,
+      "loss": 1.9315,
+      "step": 15440
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.5203042296438156e-05,
+      "loss": 1.9293,
+      "step": 15450
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.518505832923761e-05,
+      "loss": 1.9151,
+      "step": 15460
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.516706702247828e-05,
+      "loss": 1.9109,
+      "step": 15470
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.514906838485613e-05,
+      "loss": 1.8913,
+      "step": 15480
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.513106242507061e-05,
+      "loss": 1.9178,
+      "step": 15490
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.511304915182477e-05,
+      "loss": 1.9077,
+      "step": 15500
+    },
+    {
+      "epoch": 0.68,
+      "eval_loss": 1.870882272720337,
+      "eval_runtime": 11.8959,
+      "eval_samples_per_second": 344.32,
+      "eval_steps_per_second": 21.52,
+      "step": 15500
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.509502857382512e-05,
+      "loss": 1.8716,
+      "step": 15510
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.507700069978176e-05,
+      "loss": 1.9321,
+      "step": 15520
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.50589655384083e-05,
+      "loss": 1.8632,
+      "step": 15530
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.504092309842187e-05,
+      "loss": 1.8933,
+      "step": 15540
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.502287338854311e-05,
+      "loss": 1.8846,
+      "step": 15550
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.5004816417496194e-05,
+      "loss": 1.9258,
+      "step": 15560
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.4986752194008786e-05,
+      "loss": 1.9276,
+      "step": 15570
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.496868072681206e-05,
+      "loss": 1.9237,
+      "step": 15580
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.495060202464069e-05,
+      "loss": 1.9118,
+      "step": 15590
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.4932516096232864e-05,
+      "loss": 1.8694,
+      "step": 15600
+    },
+    {
+      "epoch": 0.68,
+      "eval_loss": 1.8696751594543457,
+      "eval_runtime": 11.8083,
+      "eval_samples_per_second": 346.876,
+      "eval_steps_per_second": 21.68,
+      "step": 15600
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.4914422950330247e-05,
+      "loss": 1.8977,
+      "step": 15610
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.489632259567799e-05,
+      "loss": 1.8569,
+      "step": 15620
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.487821504102474e-05,
+      "loss": 1.8957,
+      "step": 15630
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.486010029512261e-05,
+      "loss": 1.9011,
+      "step": 15640
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.48419783667272e-05,
+      "loss": 1.8822,
+      "step": 15650
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.482384926459758e-05,
+      "loss": 1.8912,
+      "step": 15660
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.480571299749628e-05,
+      "loss": 1.9014,
+      "step": 15670
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.47875695741893e-05,
+      "loss": 1.9297,
+      "step": 15680
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.476941900344611e-05,
+      "loss": 1.9134,
+      "step": 15690
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.47512612940396e-05,
+      "loss": 1.9072,
+      "step": 15700
+    },
+    {
+      "epoch": 0.69,
+      "eval_loss": 1.869808554649353,
+      "eval_runtime": 11.9339,
+      "eval_samples_per_second": 343.223,
+      "eval_steps_per_second": 21.451,
+      "step": 15700
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.473309645474614e-05,
+      "loss": 1.8311,
+      "step": 15710
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.471492449434555e-05,
+      "loss": 1.9179,
+      "step": 15720
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.4696745421621076e-05,
+      "loss": 1.8282,
+      "step": 15730
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.46785592453594e-05,
+      "loss": 1.8748,
+      "step": 15740
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.466036597435064e-05,
+      "loss": 1.8851,
+      "step": 15750
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.464216561738836e-05,
+      "loss": 1.9093,
+      "step": 15760
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.462395818326953e-05,
+      "loss": 1.874,
+      "step": 15770
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.460574368079454e-05,
+      "loss": 1.9358,
+      "step": 15780
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.458752211876721e-05,
+      "loss": 1.8965,
+      "step": 15790
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.456929350599476e-05,
+      "loss": 1.9137,
+      "step": 15800
+    },
+    {
+      "epoch": 0.69,
+      "eval_loss": 1.8691773414611816,
+      "eval_runtime": 11.9836,
+      "eval_samples_per_second": 341.799,
+      "eval_steps_per_second": 21.362,
+      "step": 15800
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.4551057851287834e-05,
+      "loss": 1.8851,
+      "step": 15810
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.4532815163460455e-05,
+      "loss": 1.8815,
+      "step": 15820
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.451456545133007e-05,
+      "loss": 1.8997,
+      "step": 15830
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.44963087237175e-05,
+      "loss": 1.8841,
+      "step": 15840
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.4478044989446965e-05,
+      "loss": 1.9042,
+      "step": 15850
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.445977425734609e-05,
+      "loss": 1.8973,
+      "step": 15860
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.444149653624585e-05,
+      "loss": 1.8587,
+      "step": 15870
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.4423211834980626e-05,
+      "loss": 1.8835,
+      "step": 15880
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.440492016238815e-05,
+      "loss": 1.9072,
+      "step": 15890
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.438662152730954e-05,
+      "loss": 1.9272,
+      "step": 15900
+    },
+    {
+      "epoch": 0.7,
+      "eval_loss": 1.8685188293457031,
+      "eval_runtime": 11.8174,
+      "eval_samples_per_second": 346.607,
+      "eval_steps_per_second": 21.663,
+      "step": 15900
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.436831593858928e-05,
+      "loss": 1.9128,
+      "step": 15910
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.435000340507519e-05,
+      "loss": 1.8712,
+      "step": 15920
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.433168393561849e-05,
+      "loss": 1.915,
+      "step": 15930
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.43133575390737e-05,
+      "loss": 1.9015,
+      "step": 15940
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.429502422429874e-05,
+      "loss": 1.8909,
+      "step": 15950
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.427668400015483e-05,
+      "loss": 1.8679,
+      "step": 15960
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.4258336875506564e-05,
+      "loss": 1.858,
+      "step": 15970
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.423998285922185e-05,
+      "loss": 1.899,
+      "step": 15980
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.422162196017194e-05,
+      "loss": 1.8772,
+      "step": 15990
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.42032541872314e-05,
+      "loss": 1.9077,
+      "step": 16000
+    },
+    {
+      "epoch": 0.7,
+      "eval_loss": 1.8670729398727417,
+      "eval_runtime": 11.9299,
+      "eval_samples_per_second": 343.34,
+      "eval_steps_per_second": 21.459,
+      "step": 16000
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.418487954927812e-05,
+      "loss": 1.9019,
+      "step": 16010
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.416649805519333e-05,
+      "loss": 1.8496,
+      "step": 16020
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.4148109713861536e-05,
+      "loss": 1.8713,
+      "step": 16030
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.412971453417059e-05,
+      "loss": 1.9179,
+      "step": 16040
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.411131252501161e-05,
+      "loss": 1.9228,
+      "step": 16050
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.4092903695279074e-05,
+      "loss": 1.9031,
+      "step": 16060
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.40744880538707e-05,
+      "loss": 1.8563,
+      "step": 16070
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.405606560968752e-05,
+      "loss": 1.8679,
+      "step": 16080
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.403763637163385e-05,
+      "loss": 1.8741,
+      "step": 16090
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.401920034861731e-05,
+      "loss": 1.8663,
+      "step": 16100
+    },
+    {
+      "epoch": 0.7,
+      "eval_loss": 1.8673057556152344,
+      "eval_runtime": 12.4069,
+      "eval_samples_per_second": 330.14,
+      "eval_steps_per_second": 20.634,
+      "step": 16100
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.400075754954877e-05,
+      "loss": 1.9126,
+      "step": 16110
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.398230798334238e-05,
+      "loss": 1.8892,
+      "step": 16120
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.396385165891559e-05,
+      "loss": 1.9086,
+      "step": 16130
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.394538858518907e-05,
+      "loss": 1.8721,
+      "step": 16140
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.392691877108681e-05,
+      "loss": 1.9023,
+      "step": 16150
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.390844222553599e-05,
+      "loss": 1.905,
+      "step": 16160
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.3889958957467104e-05,
+      "loss": 1.8876,
+      "step": 16170
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.387146897581386e-05,
+      "loss": 1.8468,
+      "step": 16180
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.3852972289513224e-05,
+      "loss": 1.89,
+      "step": 16190
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.3834468907505395e-05,
+      "loss": 1.8819,
+      "step": 16200
+    },
+    {
+      "epoch": 0.71,
+      "eval_loss": 1.8665714263916016,
+      "eval_runtime": 11.567,
+      "eval_samples_per_second": 354.112,
+      "eval_steps_per_second": 22.132,
+      "step": 16200
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.3815958838733814e-05,
+      "loss": 1.8478,
+      "step": 16210
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.379744209214517e-05,
+      "loss": 1.8759,
+      "step": 16220
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.3778918676689334e-05,
+      "loss": 1.8713,
+      "step": 16230
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.376038860131945e-05,
+      "loss": 1.9362,
+      "step": 16240
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.374185187499186e-05,
+      "loss": 1.904,
+      "step": 16250
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.372330850666611e-05,
+      "loss": 1.8876,
+      "step": 16260
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.3704758505304966e-05,
+      "loss": 1.9085,
+      "step": 16270
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.368620187987442e-05,
+      "loss": 1.907,
+      "step": 16280
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.3667638639343625e-05,
+      "loss": 1.8927,
+      "step": 16290
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.364906879268495e-05,
+      "loss": 1.8798,
+      "step": 16300
+    },
+    {
+      "epoch": 0.71,
+      "eval_loss": 1.8657153844833374,
+      "eval_runtime": 11.7422,
+      "eval_samples_per_second": 348.828,
+      "eval_steps_per_second": 21.802,
+      "step": 16300
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.363049234887399e-05,
+      "loss": 1.9308,
+      "step": 16310
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.361190931688947e-05,
+      "loss": 1.8721,
+      "step": 16320
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.359331970571335e-05,
+      "loss": 1.9018,
+      "step": 16330
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.3574723524330726e-05,
+      "loss": 1.8691,
+      "step": 16340
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.355612078172991e-05,
+      "loss": 1.9025,
+      "step": 16350
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.353751148690236e-05,
+      "loss": 1.8992,
+      "step": 16360
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.35188956488427e-05,
+      "loss": 1.8235,
+      "step": 16370
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.350027327654874e-05,
+      "loss": 1.92,
+      "step": 16380
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.348164437902143e-05,
+      "loss": 1.8775,
+      "step": 16390
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.3463008965264864e-05,
+      "loss": 1.9177,
+      "step": 16400
+    },
+    {
+      "epoch": 0.72,
+      "eval_loss": 1.8651164770126343,
+      "eval_runtime": 11.6918,
+      "eval_samples_per_second": 350.33,
+      "eval_steps_per_second": 21.896,
+      "step": 16400
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.3444367044286315e-05,
+      "loss": 1.9275,
+      "step": 16410
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.3425718625096176e-05,
+      "loss": 1.8558,
+      "step": 16420
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.340706371670799e-05,
+      "loss": 1.8686,
+      "step": 16430
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.3388402328138434e-05,
+      "loss": 1.8789,
+      "step": 16440
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.336973446840733e-05,
+      "loss": 1.8479,
+      "step": 16450
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.335106014653759e-05,
+      "loss": 1.904,
+      "step": 16460
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.333237937155531e-05,
+      "loss": 1.9069,
+      "step": 16470
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.331369215248965e-05,
+      "loss": 1.9015,
+      "step": 16480
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.329499849837293e-05,
+      "loss": 1.8867,
+      "step": 16490
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.3276298418240514e-05,
+      "loss": 1.8883,
+      "step": 16500
+    },
+    {
+      "epoch": 0.72,
+      "eval_loss": 1.864149808883667,
+      "eval_runtime": 11.9658,
+      "eval_samples_per_second": 342.309,
+      "eval_steps_per_second": 21.394,
+      "step": 16500
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.325759192113095e-05,
+      "loss": 1.8479,
+      "step": 16510
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.323887901608584e-05,
+      "loss": 1.8701,
+      "step": 16520
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.3220159712149894e-05,
+      "loss": 1.8689,
+      "step": 16530
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.320143401837092e-05,
+      "loss": 1.8919,
+      "step": 16540
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.3182701943799806e-05,
+      "loss": 1.8699,
+      "step": 16550
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.316396349749054e-05,
+      "loss": 1.8623,
+      "step": 16560
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.314521868850016e-05,
+      "loss": 1.9231,
+      "step": 16570
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.312646752588881e-05,
+      "loss": 1.8877,
+      "step": 16580
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.310771001871969e-05,
+      "loss": 1.861,
+      "step": 16590
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.308894617605907e-05,
+      "loss": 1.8961,
+      "step": 16600
+    },
+    {
+      "epoch": 0.73,
+      "eval_loss": 1.8638092279434204,
+      "eval_runtime": 11.7356,
+      "eval_samples_per_second": 349.022,
+      "eval_steps_per_second": 21.814,
+      "step": 16600
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.307017600697627e-05,
+      "loss": 1.9087,
+      "step": 16610
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.30513995205437e-05,
+      "loss": 1.9017,
+      "step": 16620
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.30326167258368e-05,
+      "loss": 1.8747,
+      "step": 16630
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.301382763193404e-05,
+      "loss": 1.8697,
+      "step": 16640
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.2995032247916974e-05,
+      "loss": 1.8932,
+      "step": 16650
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.297623058287017e-05,
+      "loss": 1.8577,
+      "step": 16660
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.295742264588125e-05,
+      "loss": 1.8503,
+      "step": 16670
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.2938608446040846e-05,
+      "loss": 1.8975,
+      "step": 16680
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.2919787992442646e-05,
+      "loss": 1.9002,
+      "step": 16690
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.2900961294183326e-05,
+      "loss": 1.9155,
+      "step": 16700
+    },
+    {
+      "epoch": 0.73,
+      "eval_loss": 1.8630321025848389,
+      "eval_runtime": 11.6984,
+      "eval_samples_per_second": 350.132,
+      "eval_steps_per_second": 21.883,
+      "step": 16700
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.2882128360362616e-05,
+      "loss": 1.9088,
+      "step": 16710
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.2863289200083226e-05,
+      "loss": 1.9019,
+      "step": 16720
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.2844443822450896e-05,
+      "loss": 1.9395,
+      "step": 16730
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.282559223657437e-05,
+      "loss": 1.8582,
+      "step": 16740
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.2806734451565385e-05,
+      "loss": 1.9054,
+      "step": 16750
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.2787870476538685e-05,
+      "loss": 1.9036,
+      "step": 16760
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.276900032061198e-05,
+      "loss": 1.8571,
+      "step": 16770
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.275012399290602e-05,
+      "loss": 1.898,
+      "step": 16780
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.273124150254447e-05,
+      "loss": 1.905,
+      "step": 16790
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.271235285865404e-05,
+      "loss": 1.8843,
+      "step": 16800
+    },
+    {
+      "epoch": 0.74,
+      "eval_loss": 1.8624677658081055,
+      "eval_runtime": 11.5664,
+      "eval_samples_per_second": 354.13,
+      "eval_steps_per_second": 22.133,
+      "step": 16800
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.269345807036436e-05,
+      "loss": 1.8558,
+      "step": 16810
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.267455714680807e-05,
+      "loss": 1.9309,
+      "step": 16820
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.2655650097120746e-05,
+      "loss": 1.885,
+      "step": 16830
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.2636736930440935e-05,
+      "loss": 1.8901,
+      "step": 16840
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.261781765591016e-05,
+      "loss": 1.9033,
+      "step": 16850
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.259889228267285e-05,
+      "loss": 1.8814,
+      "step": 16860
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.257996081987644e-05,
+      "loss": 1.8316,
+      "step": 16870
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.256102327667127e-05,
+      "loss": 1.8846,
+      "step": 16880
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.254207966221062e-05,
+      "loss": 1.8539,
+      "step": 16890
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.2523129985650715e-05,
+      "loss": 1.871,
+      "step": 16900
+    },
+    {
+      "epoch": 0.74,
+      "eval_loss": 1.8623075485229492,
+      "eval_runtime": 11.6133,
+      "eval_samples_per_second": 352.699,
+      "eval_steps_per_second": 22.044,
+      "step": 16900
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.250417425615071e-05,
+      "loss": 1.8487,
+      "step": 16910
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.248521248287269e-05,
+      "loss": 1.9164,
+      "step": 16920
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.2466244674981633e-05,
+      "loss": 1.8555,
+      "step": 16930
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.2447270841645486e-05,
+      "loss": 1.8263,
+      "step": 16940
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.2428290992035055e-05,
+      "loss": 1.8682,
+      "step": 16950
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.2409305135324085e-05,
+      "loss": 1.9171,
+      "step": 16960
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.2390313280689204e-05,
+      "loss": 1.9025,
+      "step": 16970
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.237131543730997e-05,
+      "loss": 1.885,
+      "step": 16980
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.23523116143688e-05,
+      "loss": 1.9243,
+      "step": 16990
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.2333301821051024e-05,
+      "loss": 1.8804,
+      "step": 17000
+    },
+    {
+      "epoch": 0.74,
+      "eval_loss": 1.8615334033966064,
+      "eval_runtime": 11.6378,
+      "eval_samples_per_second": 351.957,
+      "eval_steps_per_second": 21.997,
+      "step": 17000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.231428606654486e-05,
+      "loss": 1.8375,
+      "step": 17010
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.229526436004138e-05,
+      "loss": 1.863,
+      "step": 17020
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.2276236710734564e-05,
+      "loss": 1.8763,
+      "step": 17030
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.2257203127821243e-05,
+      "loss": 1.9255,
+      "step": 17040
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.2238163620501145e-05,
+      "loss": 1.8753,
+      "step": 17050
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.2219118197976814e-05,
+      "loss": 1.8752,
+      "step": 17060
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.22000668694537e-05,
+      "loss": 1.8979,
+      "step": 17070
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.218100964414009e-05,
+      "loss": 1.8533,
+      "step": 17080
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.2161946531247104e-05,
+      "loss": 1.8929,
+      "step": 17090
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.214287753998873e-05,
+      "loss": 1.8709,
+      "step": 17100
+    },
+    {
+      "epoch": 0.75,
+      "eval_loss": 1.8611003160476685,
+      "eval_runtime": 11.6511,
+      "eval_samples_per_second": 351.556,
+      "eval_steps_per_second": 21.972,
+      "step": 17100
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.212380267958179e-05,
+      "loss": 1.876,
+      "step": 17110
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.210472195924595e-05,
+      "loss": 1.8903,
+      "step": 17120
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.208563538820368e-05,
+      "loss": 1.8769,
+      "step": 17130
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.206654297568033e-05,
+      "loss": 1.8677,
+      "step": 17140
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.204744473090401e-05,
+      "loss": 1.8979,
+      "step": 17150
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.2028340663105714e-05,
+      "loss": 1.904,
+      "step": 17160
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.200923078151919e-05,
+      "loss": 1.9036,
+      "step": 17170
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.199011509538104e-05,
+      "loss": 1.8592,
+      "step": 17180
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.1970993613930655e-05,
+      "loss": 1.8772,
+      "step": 17190
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.1951866346410225e-05,
+      "loss": 1.8779,
+      "step": 17200
+    },
+    {
+      "epoch": 0.75,
+      "eval_loss": 1.8605446815490723,
+      "eval_runtime": 11.5509,
+      "eval_samples_per_second": 354.603,
+      "eval_steps_per_second": 22.163,
+      "step": 17200
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.1932733302064745e-05,
+      "loss": 1.8491,
+      "step": 17210
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.191359449014197e-05,
+      "loss": 1.8851,
+      "step": 17220
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.189444991989251e-05,
+      "loss": 1.8727,
+      "step": 17230
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.187529960056969e-05,
+      "loss": 1.8844,
+      "step": 17240
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.185614354142965e-05,
+      "loss": 1.9146,
+      "step": 17250
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.1836981751731286e-05,
+      "loss": 1.8736,
+      "step": 17260
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.1817814240736294e-05,
+      "loss": 1.852,
+      "step": 17270
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.179864101770911e-05,
+      "loss": 1.9035,
+      "step": 17280
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.177946209191691e-05,
+      "loss": 1.8563,
+      "step": 17290
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.176027747262968e-05,
+      "loss": 1.9006,
+      "step": 17300
+    },
+    {
+      "epoch": 0.76,
+      "eval_loss": 1.8597712516784668,
+      "eval_runtime": 11.6159,
+      "eval_samples_per_second": 352.621,
+      "eval_steps_per_second": 22.039,
+      "step": 17300
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.1741087169120106e-05,
+      "loss": 1.9001,
+      "step": 17310
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.1721891190663674e-05,
+      "loss": 1.8468,
+      "step": 17320
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.170268954653856e-05,
+      "loss": 1.8753,
+      "step": 17330
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.1683482246025726e-05,
+      "loss": 1.8843,
+      "step": 17340
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.166426929840883e-05,
+      "loss": 1.8496,
+      "step": 17350
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.1645050712974264e-05,
+      "loss": 1.859,
+      "step": 17360
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.162582649901118e-05,
+      "loss": 1.8695,
+      "step": 17370
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.16065966658114e-05,
+      "loss": 1.9075,
+      "step": 17380
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.1587361222669506e-05,
+      "loss": 1.8326,
+      "step": 17390
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.156812017888276e-05,
+      "loss": 1.8529,
+      "step": 17400
+    },
+    {
+      "epoch": 0.76,
+      "eval_loss": 1.8598828315734863,
+      "eval_runtime": 11.5771,
+      "eval_samples_per_second": 353.803,
+      "eval_steps_per_second": 22.113,
+      "step": 17400
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.154887354375116e-05,
+      "loss": 1.8753,
+      "step": 17410
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.1529621326577375e-05,
+      "loss": 1.8727,
+      "step": 17420
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.1510363536666794e-05,
+      "loss": 1.8874,
+      "step": 17430
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.14911001833275e-05,
+      "loss": 1.8964,
+      "step": 17440
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.147183127587026e-05,
+      "loss": 1.9168,
+      "step": 17450
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.14525568236085e-05,
+      "loss": 1.8852,
+      "step": 17460
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.143327683585837e-05,
+      "loss": 1.8632,
+      "step": 17470
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.141399132193867e-05,
+      "loss": 1.8671,
+      "step": 17480
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.1394700291170874e-05,
+      "loss": 1.8999,
+      "step": 17490
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.1375403752879135e-05,
+      "loss": 1.9191,
+      "step": 17500
+    },
+    {
+      "epoch": 0.77,
+      "eval_loss": 1.85947585105896,
+      "eval_runtime": 11.8448,
+      "eval_samples_per_second": 345.805,
+      "eval_steps_per_second": 21.613,
+      "step": 17500
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.135610171639025e-05,
+      "loss": 1.8976,
+      "step": 17510
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.133679419103368e-05,
+      "loss": 1.8638,
+      "step": 17520
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.1317481186141555e-05,
+      "loss": 1.8592,
+      "step": 17530
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.129816271104861e-05,
+      "loss": 1.9085,
+      "step": 17540
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.1278838775092277e-05,
+      "loss": 1.9276,
+      "step": 17550
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.125950938761259e-05,
+      "loss": 1.914,
+      "step": 17560
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.1240174557952245e-05,
+      "loss": 1.8823,
+      "step": 17570
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.122083429545655e-05,
+      "loss": 1.8713,
+      "step": 17580
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.120148860947343e-05,
+      "loss": 1.9081,
+      "step": 17590
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.118213750935346e-05,
+      "loss": 1.8539,
+      "step": 17600
+    },
+    {
+      "epoch": 0.77,
+      "eval_loss": 1.8588768243789673,
+      "eval_runtime": 11.5896,
+      "eval_samples_per_second": 353.419,
+      "eval_steps_per_second": 22.089,
+      "step": 17600
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.1162781004449816e-05,
+      "loss": 1.875,
+      "step": 17610
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.114341910411829e-05,
+      "loss": 1.8625,
+      "step": 17620
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.112405181771726e-05,
+      "loss": 1.9175,
+      "step": 17630
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.110467915460775e-05,
+      "loss": 1.9035,
+      "step": 17640
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.108530112415334e-05,
+      "loss": 1.8611,
+      "step": 17650
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.106591773572023e-05,
+      "loss": 1.8985,
+      "step": 17660
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.104652899867721e-05,
+      "loss": 1.9136,
+      "step": 17670
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.1027134922395656e-05,
+      "loss": 1.9122,
+      "step": 17680
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.1007735516249484e-05,
+      "loss": 1.8883,
+      "step": 17690
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.098833078961526e-05,
+      "loss": 1.8788,
+      "step": 17700
+    },
+    {
+      "epoch": 0.77,
+      "eval_loss": 1.8577911853790283,
+      "eval_runtime": 11.7168,
+      "eval_samples_per_second": 349.584,
+      "eval_steps_per_second": 21.849,
+      "step": 17700
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.0968920751872036e-05,
+      "loss": 1.8542,
+      "step": 17710
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.0949505412401516e-05,
+      "loss": 1.9247,
+      "step": 17720
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.0930084780587914e-05,
+      "loss": 1.8587,
+      "step": 17730
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.0910658865817996e-05,
+      "loss": 1.8966,
+      "step": 17740
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.089122767748113e-05,
+      "loss": 1.8524,
+      "step": 17750
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.087179122496918e-05,
+      "loss": 1.836,
+      "step": 17760
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.085234951767658e-05,
+      "loss": 1.8473,
+      "step": 17770
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.083290256500031e-05,
+      "loss": 1.8628,
+      "step": 17780
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.081345037633988e-05,
+      "loss": 1.861,
+      "step": 17790
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.079399296109731e-05,
+      "loss": 1.9114,
+      "step": 17800
+    },
+    {
+      "epoch": 0.78,
+      "eval_loss": 1.8577276468276978,
+      "eval_runtime": 11.5497,
+      "eval_samples_per_second": 354.642,
+      "eval_steps_per_second": 22.165,
+      "step": 17800
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.077453032867717e-05,
+      "loss": 1.8936,
+      "step": 17810
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.075506248848656e-05,
+      "loss": 1.8647,
+      "step": 17820
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.073558944993506e-05,
+      "loss": 1.9102,
+      "step": 17830
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.07161112224348e-05,
+      "loss": 1.9068,
+      "step": 17840
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.0696627815400386e-05,
+      "loss": 1.8272,
+      "step": 17850
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.0677139238248966e-05,
+      "loss": 1.8715,
+      "step": 17860
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.0657645500400155e-05,
+      "loss": 1.9281,
+      "step": 17870
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.063814661127607e-05,
+      "loss": 1.9073,
+      "step": 17880
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.061864258030132e-05,
+      "loss": 1.8984,
+      "step": 17890
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.0599133416903e-05,
+      "loss": 1.8766,
+      "step": 17900
+    },
+    {
+      "epoch": 0.78,
+      "eval_loss": 1.8569570779800415,
+      "eval_runtime": 11.5288,
+      "eval_samples_per_second": 355.285,
+      "eval_steps_per_second": 22.205,
+      "step": 17900
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.05796191305107e-05,
+      "loss": 1.8893,
+      "step": 17910
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.056009973055645e-05,
+      "loss": 1.8882,
+      "step": 17920
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.0540575226474785e-05,
+      "loss": 1.8622,
+      "step": 17930
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.052104562770269e-05,
+      "loss": 1.9042,
+      "step": 17940
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.0501510943679616e-05,
+      "loss": 1.8958,
+      "step": 17950
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.0481971183847475e-05,
+      "loss": 1.9045,
+      "step": 17960
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.046242635765064e-05,
+      "loss": 1.906,
+      "step": 17970
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.044287647453592e-05,
+      "loss": 1.8578,
+      "step": 17980
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.042332154395256e-05,
+      "loss": 1.8961,
+      "step": 17990
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.040376157535226e-05,
+      "loss": 1.8905,
+      "step": 18000
+    },
+    {
+      "epoch": 0.79,
+      "eval_loss": 1.8565237522125244,
+      "eval_runtime": 11.8383,
+      "eval_samples_per_second": 345.996,
+      "eval_steps_per_second": 21.625,
+      "step": 18000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.038419657818916e-05,
+      "loss": 1.8752,
+      "step": 18010
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.036462656191983e-05,
+      "loss": 1.847,
+      "step": 18020
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.0345051536003235e-05,
+      "loss": 1.8536,
+      "step": 18030
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.03254715099008e-05,
+      "loss": 1.9066,
+      "step": 18040
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.0305886493076335e-05,
+      "loss": 1.913,
+      "step": 18050
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.028629649499611e-05,
+      "loss": 1.8413,
+      "step": 18060
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.026670152512874e-05,
+      "loss": 1.9036,
+      "step": 18070
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.024710159294529e-05,
+      "loss": 1.8875,
+      "step": 18080
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.02274967079192e-05,
+      "loss": 1.866,
+      "step": 18090
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.020788687952632e-05,
+      "loss": 1.9219,
+      "step": 18100
+    },
+    {
+      "epoch": 0.79,
+      "eval_loss": 1.8559688329696655,
+      "eval_runtime": 11.6536,
+      "eval_samples_per_second": 351.479,
+      "eval_steps_per_second": 21.967,
+      "step": 18100
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.018827211724487e-05,
+      "loss": 1.8589,
+      "step": 18110
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.016865243055546e-05,
+      "loss": 1.9037,
+      "step": 18120
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.0149027828941115e-05,
+      "loss": 1.8967,
+      "step": 18130
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.012939832188718e-05,
+      "loss": 1.924,
+      "step": 18140
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.0109763918881405e-05,
+      "loss": 1.8552,
+      "step": 18150
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.00901246294139e-05,
+      "loss": 1.8851,
+      "step": 18160
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.007048046297714e-05,
+      "loss": 1.8228,
+      "step": 18170
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.005083142906594e-05,
+      "loss": 1.8643,
+      "step": 18180
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.003117753717749e-05,
+      "loss": 1.8684,
+      "step": 18190
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.001151879681132e-05,
+      "loss": 1.9042,
+      "step": 18200
+    },
+    {
+      "epoch": 0.8,
+      "eval_loss": 1.8557652235031128,
+      "eval_runtime": 11.5663,
+      "eval_samples_per_second": 354.133,
+      "eval_steps_per_second": 22.133,
+      "step": 18200
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.999185521746929e-05,
+      "loss": 1.8764,
+      "step": 18210
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.9972186808655624e-05,
+      "loss": 1.8506,
+      "step": 18220
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.9952513579876855e-05,
+      "loss": 1.8802,
+      "step": 18230
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.993283554064187e-05,
+      "loss": 1.8562,
+      "step": 18240
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.9913152700461876e-05,
+      "loss": 1.9276,
+      "step": 18250
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.9893465068850366e-05,
+      "loss": 1.8899,
+      "step": 18260
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.98737726553232e-05,
+      "loss": 1.8703,
+      "step": 18270
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.9854075469398514e-05,
+      "loss": 1.9109,
+      "step": 18280
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.983437352059677e-05,
+      "loss": 1.8824,
+      "step": 18290
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.981466681844071e-05,
+      "loss": 1.8573,
+      "step": 18300
+    },
+    {
+      "epoch": 0.8,
+      "eval_loss": 1.854828119277954,
+      "eval_runtime": 11.5928,
+      "eval_samples_per_second": 353.322,
+      "eval_steps_per_second": 22.083,
+      "step": 18300
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.97949553724554e-05,
+      "loss": 1.8958,
+      "step": 18310
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.977523919216819e-05,
+      "loss": 1.9016,
+      "step": 18320
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.975551828710871e-05,
+      "loss": 1.8862,
+      "step": 18330
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.973579266680888e-05,
+      "loss": 1.8684,
+      "step": 18340
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.971606234080289e-05,
+      "loss": 1.8834,
+      "step": 18350
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.969632731862722e-05,
+      "loss": 1.8579,
+      "step": 18360
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.967658760982061e-05,
+      "loss": 1.905,
+      "step": 18370
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.965684322392407e-05,
+      "loss": 1.8533,
+      "step": 18380
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.963709417048087e-05,
+      "loss": 1.8636,
+      "step": 18390
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.961734045903652e-05,
+      "loss": 1.8764,
+      "step": 18400
+    },
+    {
+      "epoch": 0.81,
+      "eval_loss": 1.8546396493911743,
+      "eval_runtime": 11.6105,
+      "eval_samples_per_second": 352.785,
+      "eval_steps_per_second": 22.049,
+      "step": 18400
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.959758209913881e-05,
+      "loss": 1.9181,
+      "step": 18410
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.957781910033776e-05,
+      "loss": 1.8662,
+      "step": 18420
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.955805147218563e-05,
+      "loss": 1.8574,
+      "step": 18430
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.953827922423692e-05,
+      "loss": 1.8516,
+      "step": 18440
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.9518502366048375e-05,
+      "loss": 1.8576,
+      "step": 18450
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.949872090717894e-05,
+      "loss": 1.8712,
+      "step": 18460
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.947893485718982e-05,
+      "loss": 1.8648,
+      "step": 18470
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.945914422564441e-05,
+      "loss": 1.9053,
+      "step": 18480
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.943934902210834e-05,
+      "loss": 1.8801,
+      "step": 18490
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.941954925614943e-05,
+      "loss": 1.8713,
+      "step": 18500
+    },
+    {
+      "epoch": 0.81,
+      "eval_loss": 1.854550838470459,
+      "eval_runtime": 11.781,
+      "eval_samples_per_second": 347.68,
+      "eval_steps_per_second": 21.73,
+      "step": 18500
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.939974493733773e-05,
+      "loss": 1.8499,
+      "step": 18510
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.9379936075245464e-05,
+      "loss": 1.8369,
+      "step": 18520
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.9360122679447055e-05,
+      "loss": 1.8367,
+      "step": 18530
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.934030475951915e-05,
+      "loss": 1.8849,
+      "step": 18540
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.932048232504053e-05,
+      "loss": 1.8544,
+      "step": 18550
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.930065538559222e-05,
+      "loss": 1.9585,
+      "step": 18560
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.928082395075736e-05,
+      "loss": 1.8535,
+      "step": 18570
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.926098803012132e-05,
+      "loss": 1.8866,
+      "step": 18580
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.924114763327159e-05,
+      "loss": 1.8637,
+      "step": 18590
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.9221302769797836e-05,
+      "loss": 1.8889,
+      "step": 18600
+    },
+    {
+      "epoch": 0.81,
+      "eval_loss": 1.8540050983428955,
+      "eval_runtime": 11.6301,
+      "eval_samples_per_second": 352.189,
+      "eval_steps_per_second": 22.012,
+      "step": 18600
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.920145344929192e-05,
+      "loss": 1.8645,
+      "step": 18610
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.9181599681347816e-05,
+      "loss": 1.8928,
+      "step": 18620
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.9161741475561654e-05,
+      "loss": 1.8368,
+      "step": 18630
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.914187884153171e-05,
+      "loss": 1.8581,
+      "step": 18640
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.912201178885843e-05,
+      "loss": 1.8422,
+      "step": 18650
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.910214032714434e-05,
+      "loss": 1.8837,
+      "step": 18660
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.9082264465994165e-05,
+      "loss": 1.8372,
+      "step": 18670
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.9062384215014696e-05,
+      "loss": 1.8619,
+      "step": 18680
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.904249958381487e-05,
+      "loss": 1.841,
+      "step": 18690
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.902261058200576e-05,
+      "loss": 1.8825,
+      "step": 18700
+    },
+    {
+      "epoch": 0.82,
+      "eval_loss": 1.8539540767669678,
+      "eval_runtime": 11.6061,
+      "eval_samples_per_second": 352.918,
+      "eval_steps_per_second": 22.057,
+      "step": 18700
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.900271721920051e-05,
+      "loss": 1.8167,
+      "step": 18710
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.8982819505014414e-05,
+      "loss": 1.8565,
+      "step": 18720
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.896291744906482e-05,
+      "loss": 1.8591,
+      "step": 18730
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.8943011060971254e-05,
+      "loss": 1.8222,
+      "step": 18740
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.8923100350355236e-05,
+      "loss": 1.8635,
+      "step": 18750
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.890318532684046e-05,
+      "loss": 1.8953,
+      "step": 18760
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.888326600005264e-05,
+      "loss": 1.8781,
+      "step": 18770
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.8863342379619634e-05,
+      "loss": 1.8203,
+      "step": 18780
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.884341447517132e-05,
+      "loss": 1.8809,
+      "step": 18790
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.882348229633967e-05,
+      "loss": 1.8686,
+      "step": 18800
+    },
+    {
+      "epoch": 0.82,
+      "eval_loss": 1.8534257411956787,
+      "eval_runtime": 11.5432,
+      "eval_samples_per_second": 354.841,
+      "eval_steps_per_second": 22.178,
+      "step": 18800
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.8803545852758726e-05,
+      "loss": 1.8958,
+      "step": 18810
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.8783605154064566e-05,
+      "loss": 1.8794,
+      "step": 18820
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.8763660209895374e-05,
+      "loss": 1.9037,
+      "step": 18830
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.8743711029891335e-05,
+      "loss": 1.9141,
+      "step": 18840
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.872375762369471e-05,
+      "loss": 1.8257,
+      "step": 18850
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.87038000009498e-05,
+      "loss": 1.8494,
+      "step": 18860
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.8683838171302935e-05,
+      "loss": 1.8758,
+      "step": 18870
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.8663872144402466e-05,
+      "loss": 1.8392,
+      "step": 18880
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.864390192989881e-05,
+      "loss": 1.8782,
+      "step": 18890
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.862392753744438e-05,
+      "loss": 1.8514,
+      "step": 18900
+    },
+    {
+      "epoch": 0.83,
+      "eval_loss": 1.8528721332550049,
+      "eval_runtime": 11.544,
+      "eval_samples_per_second": 354.817,
+      "eval_steps_per_second": 22.176,
+      "step": 18900
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.860394897669361e-05,
+      "loss": 1.8893,
+      "step": 18910
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.858396625730297e-05,
+      "loss": 1.9059,
+      "step": 18920
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.8563979388930914e-05,
+      "loss": 1.8665,
+      "step": 18930
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.8543988381237904e-05,
+      "loss": 1.8418,
+      "step": 18940
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.852399324388642e-05,
+      "loss": 1.8945,
+      "step": 18950
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.850399398654093e-05,
+      "loss": 1.8753,
+      "step": 18960
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.848399061886789e-05,
+      "loss": 1.8824,
+      "step": 18970
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.8463983150535735e-05,
+      "loss": 1.8406,
+      "step": 18980
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.844397159121491e-05,
+      "loss": 1.8191,
+      "step": 18990
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.8423955950577806e-05,
+      "loss": 1.8585,
+      "step": 19000
+    },
+    {
+      "epoch": 0.83,
+      "eval_loss": 1.8524885177612305,
+      "eval_runtime": 11.8631,
+      "eval_samples_per_second": 345.273,
+      "eval_steps_per_second": 21.58,
+      "step": 19000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.840393623829879e-05,
+      "loss": 1.903,
+      "step": 19010
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.838391246405423e-05,
+      "loss": 1.8799,
+      "step": 19020
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.8363884637522414e-05,
+      "loss": 1.9105,
+      "step": 19030
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.834385276838362e-05,
+      "loss": 1.8475,
+      "step": 19040
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.832381686632006e-05,
+      "loss": 1.9121,
+      "step": 19050
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.830377694101592e-05,
+      "loss": 1.9089,
+      "step": 19060
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.8283733002157296e-05,
+      "loss": 1.8896,
+      "step": 19070
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.8263685059432246e-05,
+      "loss": 1.8935,
+      "step": 19080
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.8243633122530754e-05,
+      "loss": 1.8786,
+      "step": 19090
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.8223577201144766e-05,
+      "loss": 1.8583,
+      "step": 19100
+    },
+    {
+      "epoch": 0.84,
+      "eval_loss": 1.8515942096710205,
+      "eval_runtime": 11.739,
+      "eval_samples_per_second": 348.922,
+      "eval_steps_per_second": 21.808,
+      "step": 19100
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.82035173049681e-05,
+      "loss": 1.9034,
+      "step": 19110
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.8183453443696535e-05,
+      "loss": 1.9279,
+      "step": 19120
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.816338562702775e-05,
+      "loss": 1.937,
+      "step": 19130
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.8143313864661336e-05,
+      "loss": 1.8961,
+      "step": 19140
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.812323816629882e-05,
+      "loss": 1.8915,
+      "step": 19150
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.810315854164357e-05,
+      "loss": 1.8638,
+      "step": 19160
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.808307500040091e-05,
+      "loss": 1.8666,
+      "step": 19170
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.8062987552278034e-05,
+      "loss": 1.8889,
+      "step": 19180
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.8042896206984024e-05,
+      "loss": 1.8346,
+      "step": 19190
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.802280097422984e-05,
+      "loss": 1.8317,
+      "step": 19200
+    },
+    {
+      "epoch": 0.84,
+      "eval_loss": 1.8512599468231201,
+      "eval_runtime": 11.4851,
+      "eval_samples_per_second": 356.637,
+      "eval_steps_per_second": 22.29,
+      "step": 19200
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.800270186372836e-05,
+      "loss": 1.8443,
+      "step": 19210
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.798259888519426e-05,
+      "loss": 1.8674,
+      "step": 19220
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.796249204834416e-05,
+      "loss": 1.8872,
+      "step": 19230
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.794238136289651e-05,
+      "loss": 1.8847,
+      "step": 19240
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.792226683857162e-05,
+      "loss": 1.9199,
+      "step": 19250
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.790214848509166e-05,
+      "loss": 1.8791,
+      "step": 19260
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.788202631218066e-05,
+      "loss": 1.869,
+      "step": 19270
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.7861900329564485e-05,
+      "loss": 1.8384,
+      "step": 19280
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.784177054697083e-05,
+      "loss": 1.8943,
+      "step": 19290
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.782163697412927e-05,
+      "loss": 1.8706,
+      "step": 19300
+    },
+    {
+      "epoch": 0.84,
+      "eval_loss": 1.850408911705017,
+      "eval_runtime": 11.5994,
+      "eval_samples_per_second": 353.121,
+      "eval_steps_per_second": 22.07,
+      "step": 19300
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.780149962077115e-05,
+      "loss": 1.87,
+      "step": 19310
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.7781358496629704e-05,
+      "loss": 1.8783,
+      "step": 19320
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.776121361143995e-05,
+      "loss": 1.9046,
+      "step": 19330
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.774106497493872e-05,
+      "loss": 1.9047,
+      "step": 19340
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.772091259686469e-05,
+      "loss": 1.8444,
+      "step": 19350
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.770075648695832e-05,
+      "loss": 1.8606,
+      "step": 19360
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.7680596654961886e-05,
+      "loss": 1.8874,
+      "step": 19370
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.7660433110619447e-05,
+      "loss": 1.863,
+      "step": 19380
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.764026586367687e-05,
+      "loss": 1.8678,
+      "step": 19390
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.762009492388182e-05,
+      "loss": 1.8803,
+      "step": 19400
+    },
+    {
+      "epoch": 0.85,
+      "eval_loss": 1.8501853942871094,
+      "eval_runtime": 11.6025,
+      "eval_samples_per_second": 353.026,
+      "eval_steps_per_second": 22.064,
+      "step": 19400
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.759992030098373e-05,
+      "loss": 1.8829,
+      "step": 19410
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.757974200473382e-05,
+      "loss": 1.8918,
+      "step": 19420
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.755956004488508e-05,
+      "loss": 1.8583,
+      "step": 19430
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.753937443119228e-05,
+      "loss": 1.898,
+      "step": 19440
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.751918517341194e-05,
+      "loss": 1.8663,
+      "step": 19450
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.749899228130237e-05,
+      "loss": 1.8703,
+      "step": 19460
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.747879576462361e-05,
+      "loss": 1.8459,
+      "step": 19470
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.7458595633137464e-05,
+      "loss": 1.8908,
+      "step": 19480
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.743839189660748e-05,
+      "loss": 1.859,
+      "step": 19490
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.741818456479895e-05,
+      "loss": 1.9049,
+      "step": 19500
+    },
+    {
+      "epoch": 0.85,
+      "eval_loss": 1.849900245666504,
+      "eval_runtime": 11.8738,
+      "eval_samples_per_second": 344.962,
+      "eval_steps_per_second": 21.56,
+      "step": 19500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.739797364747889e-05,
+      "loss": 1.8687,
+      "step": 19510
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.737775915441608e-05,
+      "loss": 1.8764,
+      "step": 19520
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.735754109538101e-05,
+      "loss": 1.8692,
+      "step": 19530
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.733731948014587e-05,
+      "loss": 1.8887,
+      "step": 19540
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.731709431848462e-05,
+      "loss": 1.8812,
+      "step": 19550
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.729686562017288e-05,
+      "loss": 1.8412,
+      "step": 19560
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.727663339498804e-05,
+      "loss": 1.8897,
+      "step": 19570
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.725639765270913e-05,
+      "loss": 1.8781,
+      "step": 19580
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.7236158403116925e-05,
+      "loss": 1.8901,
+      "step": 19590
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.721591565599388e-05,
+      "loss": 1.8884,
+      "step": 19600
+    },
+    {
+      "epoch": 0.86,
+      "eval_loss": 1.8497211933135986,
+      "eval_runtime": 11.3914,
+      "eval_samples_per_second": 359.57,
+      "eval_steps_per_second": 22.473,
+      "step": 19600
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.7195669421124136e-05,
+      "loss": 1.8471,
+      "step": 19610
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.7175419708293526e-05,
+      "loss": 1.8422,
+      "step": 19620
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.715516652728956e-05,
+      "loss": 1.8919,
+      "step": 19630
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.713490988790144e-05,
+      "loss": 1.8633,
+      "step": 19640
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.711464979992001e-05,
+      "loss": 1.8695,
+      "step": 19650
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.709438627313782e-05,
+      "loss": 1.808,
+      "step": 19660
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.707411931734904e-05,
+      "loss": 1.8954,
+      "step": 19670
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.705384894234953e-05,
+      "loss": 1.882,
+      "step": 19680
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.703357515793678e-05,
+      "loss": 1.8588,
+      "step": 19690
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.701329797390994e-05,
+      "loss": 1.8431,
+      "step": 19700
+    },
+    {
+      "epoch": 0.86,
+      "eval_loss": 1.8491344451904297,
+      "eval_runtime": 11.3311,
+      "eval_samples_per_second": 361.483,
+      "eval_steps_per_second": 22.593,
+      "step": 19700
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.699301740006982e-05,
+      "loss": 1.896,
+      "step": 19710
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.69727334462188e-05,
+      "loss": 1.8863,
+      "step": 19720
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.695244612216101e-05,
+      "loss": 1.8949,
+      "step": 19730
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.693215543770209e-05,
+      "loss": 1.8252,
+      "step": 19740
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.691186140264938e-05,
+      "loss": 1.8675,
+      "step": 19750
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.689156402681181e-05,
+      "loss": 1.8894,
+      "step": 19760
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.687126331999992e-05,
+      "loss": 1.8632,
+      "step": 19770
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.68509592920259e-05,
+      "loss": 1.8709,
+      "step": 19780
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.6830651952703484e-05,
+      "loss": 1.8076,
+      "step": 19790
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.681034131184806e-05,
+      "loss": 1.8623,
+      "step": 19800
+    },
+    {
+      "epoch": 0.87,
+      "eval_loss": 1.8488744497299194,
+      "eval_runtime": 11.5142,
+      "eval_samples_per_second": 355.735,
+      "eval_steps_per_second": 22.233,
+      "step": 19800
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.679002737927658e-05,
+      "loss": 1.8551,
+      "step": 19810
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.67697101648076e-05,
+      "loss": 1.9459,
+      "step": 19820
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.6749389678261254e-05,
+      "loss": 1.8397,
+      "step": 19830
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.672906592945927e-05,
+      "loss": 1.8937,
+      "step": 19840
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.670873892822494e-05,
+      "loss": 1.8702,
+      "step": 19850
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.668840868438314e-05,
+      "loss": 1.8441,
+      "step": 19860
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.66680752077603e-05,
+      "loss": 1.834,
+      "step": 19870
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.6647738508184425e-05,
+      "loss": 1.8955,
+      "step": 19880
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.6627398595485076e-05,
+      "loss": 1.8395,
+      "step": 19890
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.660705547949335e-05,
+      "loss": 1.8728,
+      "step": 19900
+    },
+    {
+      "epoch": 0.87,
+      "eval_loss": 1.8482991456985474,
+      "eval_runtime": 11.5679,
+      "eval_samples_per_second": 354.082,
+      "eval_steps_per_second": 22.13,
+      "step": 19900
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.658670917004192e-05,
+      "loss": 1.823,
+      "step": 19910
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.656635967696498e-05,
+      "loss": 1.9088,
+      "step": 19920
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.654600701009828e-05,
+      "loss": 1.8349,
+      "step": 19930
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.652565117927907e-05,
+      "loss": 1.8668,
+      "step": 19940
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.650529219434618e-05,
+      "loss": 1.881,
+      "step": 19950
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.6484930065139946e-05,
+      "loss": 1.8751,
+      "step": 19960
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.646456480150217e-05,
+      "loss": 1.8724,
+      "step": 19970
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.644419641327628e-05,
+      "loss": 1.8408,
+      "step": 19980
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.6423824910307095e-05,
+      "loss": 1.8853,
+      "step": 19990
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.640345030244103e-05,
+      "loss": 1.9109,
+      "step": 20000
+    },
+    {
+      "epoch": 0.88,
+      "eval_loss": 1.848042368888855,
+      "eval_runtime": 11.6122,
+      "eval_samples_per_second": 352.732,
+      "eval_steps_per_second": 22.046,
+      "step": 20000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.638307259952593e-05,
+      "loss": 1.9015,
+      "step": 20010
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.636269181141122e-05,
+      "loss": 1.8893,
+      "step": 20020
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.6342307947947726e-05,
+      "loss": 1.8432,
+      "step": 20030
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.6321921018987815e-05,
+      "loss": 1.8465,
+      "step": 20040
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.6301531034385326e-05,
+      "loss": 1.8545,
+      "step": 20050
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.628113800399556e-05,
+      "loss": 1.8507,
+      "step": 20060
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.626074193767531e-05,
+      "loss": 1.8718,
+      "step": 20070
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.624034284528282e-05,
+      "loss": 1.8886,
+      "step": 20080
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.621994073667783e-05,
+      "loss": 1.898,
+      "step": 20090
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.6199535621721465e-05,
+      "loss": 1.8848,
+      "step": 20100
+    },
+    {
+      "epoch": 0.88,
+      "eval_loss": 1.8475103378295898,
+      "eval_runtime": 11.4108,
+      "eval_samples_per_second": 358.959,
+      "eval_steps_per_second": 22.435,
+      "step": 20100
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.61791275102764e-05,
+      "loss": 1.8565,
+      "step": 20110
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.615871641220667e-05,
+      "loss": 1.9083,
+      "step": 20120
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.613830233737782e-05,
+      "loss": 1.8759,
+      "step": 20130
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.61178852956568e-05,
+      "loss": 1.8727,
+      "step": 20140
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.609746529691198e-05,
+      "loss": 1.8918,
+      "step": 20150
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.6077042351013206e-05,
+      "loss": 1.8517,
+      "step": 20160
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.60566164678317e-05,
+      "loss": 1.9069,
+      "step": 20170
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.603618765724014e-05,
+      "loss": 1.8429,
+      "step": 20180
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.60157559291126e-05,
+      "loss": 1.832,
+      "step": 20190
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.599532129332457e-05,
+      "loss": 1.9338,
+      "step": 20200
+    },
+    {
+      "epoch": 0.88,
+      "eval_loss": 1.847186803817749,
+      "eval_runtime": 11.3797,
+      "eval_samples_per_second": 359.94,
+      "eval_steps_per_second": 22.496,
+      "step": 20200
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.597488375975294e-05,
+      "loss": 1.8703,
+      "step": 20210
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.5954443338276004e-05,
+      "loss": 1.8843,
+      "step": 20220
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.593400003877346e-05,
+      "loss": 1.8451,
+      "step": 20230
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.591355387112637e-05,
+      "loss": 1.8947,
+      "step": 20240
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.5893104845217235e-05,
+      "loss": 1.8208,
+      "step": 20250
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.5872652970929874e-05,
+      "loss": 1.9205,
+      "step": 20260
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.585219825814953e-05,
+      "loss": 1.8699,
+      "step": 20270
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.583174071676279e-05,
+      "loss": 1.8673,
+      "step": 20280
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.581128035665763e-05,
+      "loss": 1.8823,
+      "step": 20290
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.579081718772337e-05,
+      "loss": 1.8937,
+      "step": 20300
+    },
+    {
+      "epoch": 0.89,
+      "eval_loss": 1.8470911979675293,
+      "eval_runtime": 11.3691,
+      "eval_samples_per_second": 360.276,
+      "eval_steps_per_second": 22.517,
+      "step": 20300
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.577035121985069e-05,
+      "loss": 1.8222,
+      "step": 20310
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.5749882462931646e-05,
+      "loss": 1.8978,
+      "step": 20320
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.572941092685961e-05,
+      "loss": 1.8855,
+      "step": 20330
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.570893662152931e-05,
+      "loss": 1.8851,
+      "step": 20340
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.568845955683682e-05,
+      "loss": 1.8858,
+      "step": 20350
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.566797974267954e-05,
+      "loss": 1.8572,
+      "step": 20360
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.56474971889562e-05,
+      "loss": 1.8883,
+      "step": 20370
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.5627011905566854e-05,
+      "loss": 1.8731,
+      "step": 20380
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.560652390241288e-05,
+      "loss": 1.8558,
+      "step": 20390
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.558603318939696e-05,
+      "loss": 1.8459,
+      "step": 20400
+    },
+    {
+      "epoch": 0.89,
+      "eval_loss": 1.8470215797424316,
+      "eval_runtime": 11.6007,
+      "eval_samples_per_second": 353.084,
+      "eval_steps_per_second": 22.068,
+      "step": 20400
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.556553977642309e-05,
+      "loss": 1.8773,
+      "step": 20410
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.554504367339659e-05,
+      "loss": 1.8588,
+      "step": 20420
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.552454489022405e-05,
+      "loss": 1.8891,
+      "step": 20430
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.550404343681336e-05,
+      "loss": 1.8349,
+      "step": 20440
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.5483539323073735e-05,
+      "loss": 1.8599,
+      "step": 20450
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.546303255891563e-05,
+      "loss": 1.8299,
+      "step": 20460
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.544252315425081e-05,
+      "loss": 1.8787,
+      "step": 20470
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.542201111899231e-05,
+      "loss": 1.8508,
+      "step": 20480
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.5401496463054425e-05,
+      "loss": 1.8561,
+      "step": 20490
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.5380979196352735e-05,
+      "loss": 1.8587,
+      "step": 20500
+    },
+    {
+      "epoch": 0.9,
+      "eval_loss": 1.8461756706237793,
+      "eval_runtime": 11.7063,
+      "eval_samples_per_second": 349.898,
+      "eval_steps_per_second": 21.869,
+      "step": 20500
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.5360459328804064e-05,
+      "loss": 1.9049,
+      "step": 20510
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.533993687032652e-05,
+      "loss": 1.8235,
+      "step": 20520
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.5319411830839415e-05,
+      "loss": 1.8665,
+      "step": 20530
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.529888422026336e-05,
+      "loss": 1.9031,
+      "step": 20540
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.527835404852018e-05,
+      "loss": 1.8694,
+      "step": 20550
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.525782132553296e-05,
+      "loss": 1.8432,
+      "step": 20560
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.523728606122598e-05,
+      "loss": 1.9015,
+      "step": 20570
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.521674826552479e-05,
+      "loss": 1.8434,
+      "step": 20580
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.519620794835614e-05,
+      "loss": 1.8726,
+      "step": 20590
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.5175665119648e-05,
+      "loss": 1.9017,
+      "step": 20600
+    },
+    {
+      "epoch": 0.9,
+      "eval_loss": 1.845084309577942,
+      "eval_runtime": 11.6236,
+      "eval_samples_per_second": 352.386,
+      "eval_steps_per_second": 22.024,
+      "step": 20600
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.515511978932956e-05,
+      "loss": 1.8861,
+      "step": 20610
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.513457196733121e-05,
+      "loss": 1.8689,
+      "step": 20620
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.511402166358458e-05,
+      "loss": 1.8851,
+      "step": 20630
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.5093468888022435e-05,
+      "loss": 1.8506,
+      "step": 20640
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.5072913650578797e-05,
+      "loss": 1.9112,
+      "step": 20650
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.505235596118884e-05,
+      "loss": 1.8424,
+      "step": 20660
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.503179582978892e-05,
+      "loss": 1.8257,
+      "step": 20670
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.5011233266316616e-05,
+      "loss": 1.866,
+      "step": 20680
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.499066828071063e-05,
+      "loss": 1.8652,
+      "step": 20690
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.497010088291089e-05,
+      "loss": 1.8827,
+      "step": 20700
+    },
+    {
+      "epoch": 0.91,
+      "eval_loss": 1.8452777862548828,
+      "eval_runtime": 11.6079,
+      "eval_samples_per_second": 352.863,
+      "eval_steps_per_second": 22.054,
+      "step": 20700
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.494953108285843e-05,
+      "loss": 1.8998,
+      "step": 20710
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.492895889049549e-05,
+      "loss": 1.9052,
+      "step": 20720
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.490838431576544e-05,
+      "loss": 1.838,
+      "step": 20730
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.488780736861283e-05,
+      "loss": 1.9133,
+      "step": 20740
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.4867228058983316e-05,
+      "loss": 1.8919,
+      "step": 20750
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.484664639682373e-05,
+      "loss": 1.885,
+      "step": 20760
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.4826062392082035e-05,
+      "loss": 1.8721,
+      "step": 20770
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.4805476054707315e-05,
+      "loss": 1.8589,
+      "step": 20780
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.4784887394649785e-05,
+      "loss": 1.8758,
+      "step": 20790
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.476429642186078e-05,
+      "loss": 1.8723,
+      "step": 20800
+    },
+    {
+      "epoch": 0.91,
+      "eval_loss": 1.844846487045288,
+      "eval_runtime": 11.5952,
+      "eval_samples_per_second": 353.25,
+      "eval_steps_per_second": 22.078,
+      "step": 20800
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.4743703146292774e-05,
+      "loss": 1.8686,
+      "step": 20810
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.472310757789932e-05,
+      "loss": 1.8509,
+      "step": 20820
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.470250972663512e-05,
+      "loss": 1.8502,
+      "step": 20830
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.468190960245593e-05,
+      "loss": 1.8396,
+      "step": 20840
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.466130721531867e-05,
+      "loss": 1.8543,
+      "step": 20850
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.464070257518127e-05,
+      "loss": 1.8785,
+      "step": 20860
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.4620095692002815e-05,
+      "loss": 1.8533,
+      "step": 20870
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.4599486575743454e-05,
+      "loss": 1.8293,
+      "step": 20880
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.457887523636441e-05,
+      "loss": 1.8223,
+      "step": 20890
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.455826168382799e-05,
+      "loss": 1.8676,
+      "step": 20900
+    },
+    {
+      "epoch": 0.91,
+      "eval_loss": 1.844473123550415,
+      "eval_runtime": 11.7054,
+      "eval_samples_per_second": 349.924,
+      "eval_steps_per_second": 21.87,
+      "step": 20900
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.4537645928097554e-05,
+      "loss": 1.8801,
+      "step": 20910
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.451702797913756e-05,
+      "loss": 1.8434,
+      "step": 20920
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.449640784691349e-05,
+      "loss": 1.8943,
+      "step": 20930
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.4475785541391894e-05,
+      "loss": 1.8622,
+      "step": 20940
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.445516107254039e-05,
+      "loss": 1.8625,
+      "step": 20950
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.44345344503276e-05,
+      "loss": 1.86,
+      "step": 20960
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.441390568472323e-05,
+      "loss": 1.8783,
+      "step": 20970
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.439327478569801e-05,
+      "loss": 1.8514,
+      "step": 20980
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.437264176322369e-05,
+      "loss": 1.8293,
+      "step": 20990
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.4352006627273036e-05,
+      "loss": 1.8812,
+      "step": 21000
+    },
+    {
+      "epoch": 0.92,
+      "eval_loss": 1.844010829925537,
+      "eval_runtime": 11.8385,
+      "eval_samples_per_second": 345.989,
+      "eval_steps_per_second": 21.624,
+      "step": 21000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.433136938781988e-05,
+      "loss": 1.8436,
+      "step": 21010
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.431073005483902e-05,
+      "loss": 1.8674,
+      "step": 21020
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.42900886383063e-05,
+      "loss": 1.884,
+      "step": 21030
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.426944514819856e-05,
+      "loss": 1.8758,
+      "step": 21040
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.424879959449363e-05,
+      "loss": 1.8379,
+      "step": 21050
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.422815198717037e-05,
+      "loss": 1.8983,
+      "step": 21060
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.4207502336208586e-05,
+      "loss": 1.9012,
+      "step": 21070
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.418685065158912e-05,
+      "loss": 1.8474,
+      "step": 21080
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.416619694329376e-05,
+      "loss": 1.8605,
+      "step": 21090
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.414554122130529e-05,
+      "loss": 1.8589,
+      "step": 21100
+    },
+    {
+      "epoch": 0.92,
+      "eval_loss": 1.8435540199279785,
+      "eval_runtime": 11.7192,
+      "eval_samples_per_second": 349.512,
+      "eval_steps_per_second": 21.844,
+      "step": 21100
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.412488349560748e-05,
+      "loss": 1.9195,
+      "step": 21110
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.4104223776185025e-05,
+      "loss": 1.8987,
+      "step": 21120
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.408356207302363e-05,
+      "loss": 1.8662,
+      "step": 21130
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.4062898396109926e-05,
+      "loss": 1.8617,
+      "step": 21140
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.404223275543153e-05,
+      "loss": 1.8549,
+      "step": 21150
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.402156516097697e-05,
+      "loss": 1.8974,
+      "step": 21160
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.400089562273576e-05,
+      "loss": 1.8277,
+      "step": 21170
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.3980224150698304e-05,
+      "loss": 1.8674,
+      "step": 21180
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.3959550754856e-05,
+      "loss": 1.8959,
+      "step": 21190
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.3938875445201126e-05,
+      "loss": 1.8234,
+      "step": 21200
+    },
+    {
+      "epoch": 0.93,
+      "eval_loss": 1.8433294296264648,
+      "eval_runtime": 12.3336,
+      "eval_samples_per_second": 332.1,
+      "eval_steps_per_second": 20.756,
+      "step": 21200
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.39181982317269e-05,
+      "loss": 1.8358,
+      "step": 21210
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.3897519124427484e-05,
+      "loss": 1.9033,
+      "step": 21220
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.387683813329791e-05,
+      "loss": 1.8959,
+      "step": 21230
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.385615526833416e-05,
+      "loss": 1.8205,
+      "step": 21240
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.38354705395331e-05,
+      "loss": 1.8477,
+      "step": 21250
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.381478395689252e-05,
+      "loss": 1.8897,
+      "step": 21260
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.3794095530411064e-05,
+      "loss": 1.8679,
+      "step": 21270
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.377340527008831e-05,
+      "loss": 1.8254,
+      "step": 21280
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.37527131859247e-05,
+      "loss": 1.8897,
+      "step": 21290
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.373201928792158e-05,
+      "loss": 1.878,
+      "step": 21300
+    },
+    {
+      "epoch": 0.93,
+      "eval_loss": 1.8435399532318115,
+      "eval_runtime": 11.6036,
+      "eval_samples_per_second": 352.993,
+      "eval_steps_per_second": 22.062,
+      "step": 21300
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.371132358608114e-05,
+      "loss": 1.8616,
+      "step": 21310
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.3690626090406446e-05,
+      "loss": 1.8702,
+      "step": 21320
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.366992681090147e-05,
+      "loss": 1.8344,
+      "step": 21330
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.3649225757571e-05,
+      "loss": 1.898,
+      "step": 21340
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.362852294042071e-05,
+      "loss": 1.8452,
+      "step": 21350
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.3607818369457106e-05,
+      "loss": 1.872,
+      "step": 21360
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.3587112054687566e-05,
+      "loss": 1.8164,
+      "step": 21370
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.3566404006120296e-05,
+      "loss": 1.8258,
+      "step": 21380
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.3545694233764326e-05,
+      "loss": 1.8799,
+      "step": 21390
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.352498274762955e-05,
+      "loss": 1.8901,
+      "step": 21400
+    },
+    {
+      "epoch": 0.94,
+      "eval_loss": 1.8432936668395996,
+      "eval_runtime": 11.5934,
+      "eval_samples_per_second": 353.304,
+      "eval_steps_per_second": 22.082,
+      "step": 21400
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.3504269557726675e-05,
+      "loss": 1.8615,
+      "step": 21410
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.348355467406723e-05,
+      "loss": 1.8215,
+      "step": 21420
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.3462838106663544e-05,
+      "loss": 1.8374,
+      "step": 21430
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.344211986552881e-05,
+      "loss": 1.9267,
+      "step": 21440
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.342139996067699e-05,
+      "loss": 1.8219,
+      "step": 21450
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.340067840212285e-05,
+      "loss": 1.8697,
+      "step": 21460
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.3379955199881973e-05,
+      "loss": 1.883,
+      "step": 21470
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.3359230363970734e-05,
+      "loss": 1.8599,
+      "step": 21480
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.333850390440629e-05,
+      "loss": 1.8511,
+      "step": 21490
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.331777583120657e-05,
+      "loss": 1.8464,
+      "step": 21500
+    },
+    {
+      "epoch": 0.94,
+      "eval_loss": 1.843240737915039,
+      "eval_runtime": 11.8662,
+      "eval_samples_per_second": 345.183,
+      "eval_steps_per_second": 21.574,
+      "step": 21500
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.329704615439032e-05,
+      "loss": 1.8908,
+      "step": 21510
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.327631488397702e-05,
+      "loss": 1.9016,
+      "step": 21520
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.3255582029986956e-05,
+      "loss": 1.8827,
+      "step": 21530
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.323484760244115e-05,
+      "loss": 1.905,
+      "step": 21540
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.321411161136141e-05,
+      "loss": 1.872,
+      "step": 21550
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.319337406677029e-05,
+      "loss": 1.8769,
+      "step": 21560
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.317263497869107e-05,
+      "loss": 1.9206,
+      "step": 21570
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.315189435714781e-05,
+      "loss": 1.8788,
+      "step": 21580
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.31311522121653e-05,
+      "loss": 1.8762,
+      "step": 21590
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.311040855376908e-05,
+      "loss": 1.884,
+      "step": 21600
+    },
+    {
+      "epoch": 0.95,
+      "eval_loss": 1.8426785469055176,
+      "eval_runtime": 11.6483,
+      "eval_samples_per_second": 351.639,
+      "eval_steps_per_second": 21.977,
+      "step": 21600
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.3089663391985375e-05,
+      "loss": 1.8241,
+      "step": 21610
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.30689167368412e-05,
+      "loss": 1.8922,
+      "step": 21620
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.304816859836424e-05,
+      "loss": 1.8536,
+      "step": 21630
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.302741898658294e-05,
+      "loss": 1.8683,
+      "step": 21640
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.300666791152641e-05,
+      "loss": 1.8762,
+      "step": 21650
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.29859153832245e-05,
+      "loss": 1.8442,
+      "step": 21660
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.296516141170776e-05,
+      "loss": 1.8689,
+      "step": 21670
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.294440600700742e-05,
+      "loss": 1.8598,
+      "step": 21680
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.2923649179155436e-05,
+      "loss": 1.8641,
+      "step": 21690
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.290289093818442e-05,
+      "loss": 1.8715,
+      "step": 21700
+    },
+    {
+      "epoch": 0.95,
+      "eval_loss": 1.8424584865570068,
+      "eval_runtime": 11.6455,
+      "eval_samples_per_second": 351.723,
+      "eval_steps_per_second": 21.983,
+      "step": 21700
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.288213129412766e-05,
+      "loss": 1.84,
+      "step": 21710
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.286137025701916e-05,
+      "loss": 1.8564,
+      "step": 21720
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.2840607836893574e-05,
+      "loss": 1.8725,
+      "step": 21730
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.281984404378622e-05,
+      "loss": 1.8576,
+      "step": 21740
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.27990788877331e-05,
+      "loss": 1.882,
+      "step": 21750
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.277831237877085e-05,
+      "loss": 1.8279,
+      "step": 21760
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.275754452693677e-05,
+      "loss": 1.9028,
+      "step": 21770
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.273677534226883e-05,
+      "loss": 1.8716,
+      "step": 21780
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.27160048348056e-05,
+      "loss": 1.8502,
+      "step": 21790
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.269523301458633e-05,
+      "loss": 1.8384,
+      "step": 21800
+    },
+    {
+      "epoch": 0.95,
+      "eval_loss": 1.8420464992523193,
+      "eval_runtime": 11.6127,
+      "eval_samples_per_second": 352.716,
+      "eval_steps_per_second": 22.045,
+      "step": 21800
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.2674459891650884e-05,
+      "loss": 1.8778,
+      "step": 21810
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.265368547603977e-05,
+      "loss": 1.8726,
+      "step": 21820
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.263290977779409e-05,
+      "loss": 1.905,
+      "step": 21830
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.2612132806955604e-05,
+      "loss": 1.868,
+      "step": 21840
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.259135457356667e-05,
+      "loss": 1.8234,
+      "step": 21850
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.257057508767025e-05,
+      "loss": 1.8995,
+      "step": 21860
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.254979435930993e-05,
+      "loss": 1.8289,
+      "step": 21870
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.252901239852986e-05,
+      "loss": 1.864,
+      "step": 21880
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.2508229215374845e-05,
+      "loss": 1.8272,
+      "step": 21890
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.248744481989022e-05,
+      "loss": 1.8424,
+      "step": 21900
+    },
+    {
+      "epoch": 0.96,
+      "eval_loss": 1.8414499759674072,
+      "eval_runtime": 11.656,
+      "eval_samples_per_second": 351.406,
+      "eval_steps_per_second": 21.963,
+      "step": 21900
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.2466659222121936e-05,
+      "loss": 1.8475,
+      "step": 21910
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.2445872432116525e-05,
+      "loss": 1.895,
+      "step": 21920
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.242508445992109e-05,
+      "loss": 1.8655,
+      "step": 21930
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.2404295315583324e-05,
+      "loss": 1.8606,
+      "step": 21940
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.238350500915144e-05,
+      "loss": 1.8496,
+      "step": 21950
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.2362713550674256e-05,
+      "loss": 1.8838,
+      "step": 21960
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.234192095020113e-05,
+      "loss": 1.8587,
+      "step": 21970
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.232112721778198e-05,
+      "loss": 1.8999,
+      "step": 21980
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.230033236346725e-05,
+      "loss": 1.8584,
+      "step": 21990
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.227953639730798e-05,
+      "loss": 1.8248,
+      "step": 22000
+    },
+    {
+      "epoch": 0.96,
+      "eval_loss": 1.8413933515548706,
+      "eval_runtime": 11.8856,
+      "eval_samples_per_second": 344.619,
+      "eval_steps_per_second": 21.539,
+      "step": 22000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.225873932935567e-05,
+      "loss": 1.8195,
+      "step": 22010
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.223794116966241e-05,
+      "loss": 1.9048,
+      "step": 22020
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.22171419282808e-05,
+      "loss": 1.8697,
+      "step": 22030
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.219634161526396e-05,
+      "loss": 1.8435,
+      "step": 22040
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.2175540240665526e-05,
+      "loss": 1.9132,
+      "step": 22050
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.215473781453965e-05,
+      "loss": 1.8596,
+      "step": 22060
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.213393434694101e-05,
+      "loss": 1.8665,
+      "step": 22070
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.211312984792476e-05,
+      "loss": 1.8975,
+      "step": 22080
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.209232432754656e-05,
+      "loss": 1.8316,
+      "step": 22090
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.207151779586259e-05,
+      "loss": 1.8948,
+      "step": 22100
+    },
+    {
+      "epoch": 0.97,
+      "eval_loss": 1.8409616947174072,
+      "eval_runtime": 11.7108,
+      "eval_samples_per_second": 349.763,
+      "eval_steps_per_second": 21.86,
+      "step": 22100
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.205071026292946e-05,
+      "loss": 1.8931,
+      "step": 22110
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.202990173880433e-05,
+      "loss": 1.9053,
+      "step": 22120
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.2009092233544795e-05,
+      "loss": 1.9124,
+      "step": 22130
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.198828175720895e-05,
+      "loss": 1.8689,
+      "step": 22140
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.196747031985533e-05,
+      "loss": 1.8693,
+      "step": 22150
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.194665793154297e-05,
+      "loss": 1.8486,
+      "step": 22160
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.192584460233134e-05,
+      "loss": 1.8557,
+      "step": 22170
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.190503034228037e-05,
+      "loss": 1.8637,
+      "step": 22180
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.188421516145045e-05,
+      "loss": 1.8603,
+      "step": 22190
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.18633990699024e-05,
+      "loss": 1.8666,
+      "step": 22200
+    },
+    {
+      "epoch": 0.97,
+      "eval_loss": 1.8402700424194336,
+      "eval_runtime": 11.7872,
+      "eval_samples_per_second": 347.497,
+      "eval_steps_per_second": 21.719,
+      "step": 22200
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.184258207769749e-05,
+      "loss": 1.888,
+      "step": 22210
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.182176419489741e-05,
+      "loss": 1.8296,
+      "step": 22220
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.180094543156431e-05,
+      "loss": 1.8857,
+      "step": 22230
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.1780125797760735e-05,
+      "loss": 1.9119,
+      "step": 22240
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.175930530354968e-05,
+      "loss": 1.8893,
+      "step": 22250
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.173848395899453e-05,
+      "loss": 1.8697,
+      "step": 22260
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.171766177415909e-05,
+      "loss": 1.901,
+      "step": 22270
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.1696838759107575e-05,
+      "loss": 1.8397,
+      "step": 22280
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.1676014923904594e-05,
+      "loss": 1.884,
+      "step": 22290
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.165519027861517e-05,
+      "loss": 1.9316,
+      "step": 22300
+    },
+    {
+      "epoch": 0.98,
+      "eval_loss": 1.8402478694915771,
+      "eval_runtime": 11.775,
+      "eval_samples_per_second": 347.857,
+      "eval_steps_per_second": 21.741,
+      "step": 22300
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.163436483330469e-05,
+      "loss": 1.8686,
+      "step": 22310
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.161353859803895e-05,
+      "loss": 1.8972,
+      "step": 22320
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.1592711582884107e-05,
+      "loss": 1.8481,
+      "step": 22330
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.1571883797906726e-05,
+      "loss": 1.8679,
+      "step": 22340
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.155105525317372e-05,
+      "loss": 1.8732,
+      "step": 22350
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.1530225958752365e-05,
+      "loss": 1.8054,
+      "step": 22360
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.1509395924710334e-05,
+      "loss": 1.9027,
+      "step": 22370
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.1488565161115593e-05,
+      "loss": 1.8344,
+      "step": 22380
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.1467733678036546e-05,
+      "loss": 1.8781,
+      "step": 22390
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.1446901485541865e-05,
+      "loss": 1.8247,
+      "step": 22400
+    },
+    {
+      "epoch": 0.98,
+      "eval_loss": 1.8399176597595215,
+      "eval_runtime": 11.5771,
+      "eval_samples_per_second": 353.801,
+      "eval_steps_per_second": 22.113,
+      "step": 22400
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.142606859370063e-05,
+      "loss": 1.8256,
+      "step": 22410
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.140523501258219e-05,
+      "loss": 1.8657,
+      "step": 22420
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.1384400752256293e-05,
+      "loss": 1.8599,
+      "step": 22430
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.136356582279298e-05,
+      "loss": 1.8749,
+      "step": 22440
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.134273023426263e-05,
+      "loss": 1.8761,
+      "step": 22450
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.132189399673593e-05,
+      "loss": 1.8616,
+      "step": 22460
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.1301057120283876e-05,
+      "loss": 1.865,
+      "step": 22470
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.1280219614977786e-05,
+      "loss": 1.8759,
+      "step": 22480
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.125938149088927e-05,
+      "loss": 1.8637,
+      "step": 22490
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.123854275809027e-05,
+      "loss": 1.8637,
+      "step": 22500
+    },
+    {
+      "epoch": 0.98,
+      "eval_loss": 1.839624285697937,
+      "eval_runtime": 12.1106,
+      "eval_samples_per_second": 338.216,
+      "eval_steps_per_second": 21.139,
+      "step": 22500
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.1217703426652975e-05,
+      "loss": 1.8921,
+      "step": 22510
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.119686350664989e-05,
+      "loss": 1.8818,
+      "step": 22520
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.117602300815379e-05,
+      "loss": 1.8604,
+      "step": 22530
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.1155181941237754e-05,
+      "loss": 1.8978,
+      "step": 22540
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.113434031597512e-05,
+      "loss": 1.8593,
+      "step": 22550
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.1113498142439473e-05,
+      "loss": 1.8391,
+      "step": 22560
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.1092655430704725e-05,
+      "loss": 1.8613,
+      "step": 22570
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.1071812190844975e-05,
+      "loss": 1.8845,
+      "step": 22580
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.105096843293463e-05,
+      "loss": 1.8934,
+      "step": 22590
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.1030124167048326e-05,
+      "loss": 1.8463,
+      "step": 22600
+    },
+    {
+      "epoch": 0.99,
+      "eval_loss": 1.839435338973999,
+      "eval_runtime": 11.6439,
+      "eval_samples_per_second": 351.773,
+      "eval_steps_per_second": 21.986,
+      "step": 22600
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.100927940326095e-05,
+      "loss": 1.8468,
+      "step": 22610
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.0988434151647625e-05,
+      "loss": 1.8613,
+      "step": 22620
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.096758842228371e-05,
+      "loss": 1.8547,
+      "step": 22630
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.09467422252448e-05,
+      "loss": 1.8687,
+      "step": 22640
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.0925895570606714e-05,
+      "loss": 1.8653,
+      "step": 22650
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.0905048468445494e-05,
+      "loss": 1.8523,
+      "step": 22660
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.088420092883738e-05,
+      "loss": 1.8353,
+      "step": 22670
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.086335296185885e-05,
+      "loss": 1.8948,
+      "step": 22680
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.084250457758659e-05,
+      "loss": 1.8711,
+      "step": 22690
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.0821655786097465e-05,
+      "loss": 1.8616,
+      "step": 22700
+    },
+    {
+      "epoch": 0.99,
+      "eval_loss": 1.8392356634140015,
+      "eval_runtime": 11.5942,
+      "eval_samples_per_second": 353.279,
+      "eval_steps_per_second": 22.08,
+      "step": 22700
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.080080659746853e-05,
+      "loss": 1.8775,
+      "step": 22710
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.0779957021777084e-05,
+      "loss": 1.8427,
+      "step": 22720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.0759107069100556e-05,
+      "loss": 1.8399,
+      "step": 22730
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.0738256749516575e-05,
+      "loss": 1.8356,
+      "step": 22740
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.071740607310296e-05,
+      "loss": 1.8397,
+      "step": 22750
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.0696555049937705e-05,
+      "loss": 1.8664,
+      "step": 22760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.067570369009895e-05,
+      "loss": 1.889,
+      "step": 22770
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.0654852003665e-05,
+      "loss": 1.8426,
+      "step": 22780
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.0634000000714345e-05,
+      "loss": 1.8929,
+      "step": 22790
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.061314769132559e-05,
+      "loss": 1.9215,
+      "step": 22800
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 1.839365005493164,
+      "eval_runtime": 11.7046,
+      "eval_samples_per_second": 349.947,
+      "eval_steps_per_second": 21.872,
+      "step": 22800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.0592295085577536e-05,
+      "loss": 1.8533,
+      "step": 22810
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.0571442193549066e-05,
+      "loss": 1.8485,
+      "step": 22820
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.055058902531925e-05,
+      "loss": 1.8331,
+      "step": 22830
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.052973559096729e-05,
+      "loss": 1.8565,
+      "step": 22840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.0508881900572467e-05,
+      "loss": 1.8282,
+      "step": 22850
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.0488027964214257e-05,
+      "loss": 1.7852,
+      "step": 22860
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.046717379197219e-05,
+      "loss": 1.8608,
+      "step": 22870
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.0446319393925966e-05,
+      "loss": 1.784,
+      "step": 22880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.042546478015535e-05,
+      "loss": 1.8352,
+      "step": 22890
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.0404609960740227e-05,
+      "loss": 1.843,
+      "step": 22900
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 1.8390878438949585,
+      "eval_runtime": 11.6738,
+      "eval_samples_per_second": 350.871,
+      "eval_steps_per_second": 21.929,
+      "step": 22900
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.0383754945760583e-05,
+      "loss": 1.8813,
+      "step": 22910
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.0362899745296515e-05,
+      "loss": 1.8617,
+      "step": 22920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.034204436942818e-05,
+      "loss": 1.8365,
+      "step": 22930
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.0321188828235827e-05,
+      "loss": 1.8799,
+      "step": 22940
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.0300333131799806e-05,
+      "loss": 1.8485,
+      "step": 22950
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0279477290200505e-05,
+      "loss": 1.8404,
+      "step": 22960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0258621313518433e-05,
+      "loss": 1.8718,
+      "step": 22970
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0237765211834105e-05,
+      "loss": 1.8443,
+      "step": 22980
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0216908995228152e-05,
+      "loss": 1.8151,
+      "step": 22990
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0196052673781224e-05,
+      "loss": 1.778,
+      "step": 23000
+    },
+    {
+      "epoch": 1.01,
+      "eval_loss": 1.8393231630325317,
+      "eval_runtime": 12.9491,
+      "eval_samples_per_second": 316.316,
+      "eval_steps_per_second": 19.77,
+      "step": 23000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0175196257574026e-05,
+      "loss": 1.8567,
+      "step": 23010
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.015433975668733e-05,
+      "loss": 1.8481,
+      "step": 23020
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0133483181201915e-05,
+      "loss": 1.8392,
+      "step": 23030
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.011262654119864e-05,
+      "loss": 1.8451,
+      "step": 23040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0091769846758353e-05,
+      "loss": 1.8074,
+      "step": 23050
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0070913107961955e-05,
+      "loss": 1.8228,
+      "step": 23060
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0050056334890354e-05,
+      "loss": 1.8126,
+      "step": 23070
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.002919953762448e-05,
+      "loss": 1.816,
+      "step": 23080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0008342726245283e-05,
+      "loss": 1.8528,
+      "step": 23090
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9987485910833704e-05,
+      "loss": 1.8528,
+      "step": 23100
+    },
+    {
+      "epoch": 1.01,
+      "eval_loss": 1.8387082815170288,
+      "eval_runtime": 11.526,
+      "eval_samples_per_second": 355.37,
+      "eval_steps_per_second": 22.211,
+      "step": 23100
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9966629101470693e-05,
+      "loss": 1.8318,
+      "step": 23110
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.99457723082372e-05,
+      "loss": 1.7999,
+      "step": 23120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9924915541214165e-05,
+      "loss": 1.8028,
+      "step": 23130
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9904058810482516e-05,
+      "loss": 1.82,
+      "step": 23140
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9883202126123156e-05,
+      "loss": 1.827,
+      "step": 23150
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9862345498216987e-05,
+      "loss": 1.8541,
+      "step": 23160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9841488936844853e-05,
+      "loss": 1.8234,
+      "step": 23170
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9820632452087598e-05,
+      "loss": 1.8179,
+      "step": 23180
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9799776054025987e-05,
+      "loss": 1.7522,
+      "step": 23190
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9778919752740805e-05,
+      "loss": 1.8131,
+      "step": 23200
+    },
+    {
+      "epoch": 1.02,
+      "eval_loss": 1.8382573127746582,
+      "eval_runtime": 11.5966,
+      "eval_samples_per_second": 353.208,
+      "eval_steps_per_second": 22.075,
+      "step": 23200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9758063558312715e-05,
+      "loss": 1.8818,
+      "step": 23210
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9737207480822403e-05,
+      "loss": 1.8894,
+      "step": 23220
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9716351530350437e-05,
+      "loss": 1.8496,
+      "step": 23230
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9695495716977364e-05,
+      "loss": 1.8393,
+      "step": 23240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9674640050783645e-05,
+      "loss": 1.8029,
+      "step": 23250
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9653784541849667e-05,
+      "loss": 1.836,
+      "step": 23260
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9632929200255768e-05,
+      "loss": 1.8207,
+      "step": 23270
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.961207403608217e-05,
+      "loss": 1.8362,
+      "step": 23280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9591219059409042e-05,
+      "loss": 1.845,
+      "step": 23290
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.957036428031641e-05,
+      "loss": 1.841,
+      "step": 23300
+    },
+    {
+      "epoch": 1.02,
+      "eval_loss": 1.8381986618041992,
+      "eval_runtime": 11.5438,
+      "eval_samples_per_second": 354.822,
+      "eval_steps_per_second": 22.176,
+      "step": 23300
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9549509708884293e-05,
+      "loss": 1.7825,
+      "step": 23310
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9528655355192507e-05,
+      "loss": 1.8415,
+      "step": 23320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9507801229320857e-05,
+      "loss": 1.8185,
+      "step": 23330
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9486947341348954e-05,
+      "loss": 1.7948,
+      "step": 23340
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9466093701356362e-05,
+      "loss": 1.8576,
+      "step": 23350
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9445240319422474e-05,
+      "loss": 1.8236,
+      "step": 23360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9424387205626606e-05,
+      "loss": 1.821,
+      "step": 23370
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9403534370047895e-05,
+      "loss": 1.8555,
+      "step": 23380
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.938268182276539e-05,
+      "loss": 1.8439,
+      "step": 23390
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9361829573857957e-05,
+      "loss": 1.887,
+      "step": 23400
+    },
+    {
+      "epoch": 1.02,
+      "eval_loss": 1.8380987644195557,
+      "eval_runtime": 11.4898,
+      "eval_samples_per_second": 356.489,
+      "eval_steps_per_second": 22.281,
+      "step": 23400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.934097763340436e-05,
+      "loss": 1.8797,
+      "step": 23410
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.9320126011483177e-05,
+      "loss": 1.9105,
+      "step": 23420
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.929927471817286e-05,
+      "loss": 1.8606,
+      "step": 23430
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.9278423763551684e-05,
+      "loss": 1.8446,
+      "step": 23440
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.9257573157697777e-05,
+      "loss": 1.8375,
+      "step": 23450
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.923672291068906e-05,
+      "loss": 1.8811,
+      "step": 23460
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.9215873032603347e-05,
+      "loss": 1.812,
+      "step": 23470
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.9195023533518214e-05,
+      "loss": 1.8409,
+      "step": 23480
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.917417442351107e-05,
+      "loss": 1.8653,
+      "step": 23490
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.9153325712659155e-05,
+      "loss": 1.8448,
+      "step": 23500
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 1.8378263711929321,
+      "eval_runtime": 11.9067,
+      "eval_samples_per_second": 344.008,
+      "eval_steps_per_second": 21.501,
+      "step": 23500
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.9132477411039492e-05,
+      "loss": 1.753,
+      "step": 23510
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.9111629528728927e-05,
+      "loss": 1.816,
+      "step": 23520
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.9090782075804082e-05,
+      "loss": 1.8389,
+      "step": 23530
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.906993506234139e-05,
+      "loss": 1.7968,
+      "step": 23540
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.904908849841706e-05,
+      "loss": 1.8927,
+      "step": 23550
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.9028242394107096e-05,
+      "loss": 1.8586,
+      "step": 23560
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.900739675948725e-05,
+      "loss": 1.8503,
+      "step": 23570
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.8986551604633098e-05,
+      "loss": 1.8244,
+      "step": 23580
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.8965706939619926e-05,
+      "loss": 1.8273,
+      "step": 23590
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.8944862774522838e-05,
+      "loss": 1.8311,
+      "step": 23600
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 1.837775707244873,
+      "eval_runtime": 11.4396,
+      "eval_samples_per_second": 358.055,
+      "eval_steps_per_second": 22.378,
+      "step": 23600
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.8924019119416648e-05,
+      "loss": 1.7796,
+      "step": 23610
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.890317598437596e-05,
+      "loss": 1.8699,
+      "step": 23620
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.88823333794751e-05,
+      "loss": 1.8105,
+      "step": 23630
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.8861491314788167e-05,
+      "loss": 1.8384,
+      "step": 23640
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.884064980038896e-05,
+      "loss": 1.8276,
+      "step": 23650
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.881980884635106e-05,
+      "loss": 1.8022,
+      "step": 23660
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.8798968462747725e-05,
+      "loss": 1.7894,
+      "step": 23670
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.8778128659651985e-05,
+      "loss": 1.842,
+      "step": 23680
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.875728944713655e-05,
+      "loss": 1.8128,
+      "step": 23690
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.873645083527388e-05,
+      "loss": 1.8564,
+      "step": 23700
+    },
+    {
+      "epoch": 1.04,
+      "eval_loss": 1.8370027542114258,
+      "eval_runtime": 13.2784,
+      "eval_samples_per_second": 308.47,
+      "eval_steps_per_second": 19.279,
+      "step": 23700
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.8715612834136125e-05,
+      "loss": 1.8202,
+      "step": 23710
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.8694775453795117e-05,
+      "loss": 1.7821,
+      "step": 23720
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.867393870432244e-05,
+      "loss": 1.8326,
+      "step": 23730
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.865310259578933e-05,
+      "loss": 1.8237,
+      "step": 23740
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.863226713826674e-05,
+      "loss": 1.8495,
+      "step": 23750
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.8611432341825292e-05,
+      "loss": 1.8296,
+      "step": 23760
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.8590598216535298e-05,
+      "loss": 1.8261,
+      "step": 23770
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.8569764772466733e-05,
+      "loss": 1.834,
+      "step": 23780
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.854893201968926e-05,
+      "loss": 1.8588,
+      "step": 23790
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.8528099968272194e-05,
+      "loss": 1.8609,
+      "step": 23800
+    },
+    {
+      "epoch": 1.04,
+      "eval_loss": 1.837214708328247,
+      "eval_runtime": 11.7272,
+      "eval_samples_per_second": 349.274,
+      "eval_steps_per_second": 21.83,
+      "step": 23800
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.850726862828452e-05,
+      "loss": 1.8529,
+      "step": 23810
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.8486438009794868e-05,
+      "loss": 1.7861,
+      "step": 23820
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.8465608122871547e-05,
+      "loss": 1.8063,
+      "step": 23830
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.8444778977582455e-05,
+      "loss": 1.8232,
+      "step": 23840
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.842395058399521e-05,
+      "loss": 1.8102,
+      "step": 23850
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.840312295217699e-05,
+      "loss": 1.8094,
+      "step": 23860
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.8382296092194664e-05,
+      "loss": 1.831,
+      "step": 23870
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.836147001411469e-05,
+      "loss": 1.8275,
+      "step": 23880
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.8340644728003173e-05,
+      "loss": 1.8325,
+      "step": 23890
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.8319820243925808e-05,
+      "loss": 1.8287,
+      "step": 23900
+    },
+    {
+      "epoch": 1.05,
+      "eval_loss": 1.8369483947753906,
+      "eval_runtime": 11.5483,
+      "eval_samples_per_second": 354.685,
+      "eval_steps_per_second": 22.168,
+      "step": 23900
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.829899657194794e-05,
+      "loss": 1.8468,
+      "step": 23910
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.827817372213448e-05,
+      "loss": 1.8562,
+      "step": 23920
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.8257351704549975e-05,
+      "loss": 1.8662,
+      "step": 23930
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.8236530529258547e-05,
+      "loss": 1.8251,
+      "step": 23940
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.821571020632393e-05,
+      "loss": 1.805,
+      "step": 23950
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.8194890745809443e-05,
+      "loss": 1.802,
+      "step": 23960
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.817407215777795e-05,
+      "loss": 1.7749,
+      "step": 23970
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.8153254452291956e-05,
+      "loss": 1.7909,
+      "step": 23980
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.8132437639413486e-05,
+      "loss": 1.7694,
+      "step": 23990
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.8111621729204177e-05,
+      "loss": 1.8344,
+      "step": 24000
+    },
+    {
+      "epoch": 1.05,
+      "eval_loss": 1.8369200229644775,
+      "eval_runtime": 11.8804,
+      "eval_samples_per_second": 344.77,
+      "eval_steps_per_second": 21.548,
+      "step": 24000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.809080673172519e-05,
+      "loss": 1.7981,
+      "step": 24010
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.8069992657037273e-05,
+      "loss": 1.8076,
+      "step": 24020
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.804917951520071e-05,
+      "loss": 1.794,
+      "step": 24030
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.8028367316275357e-05,
+      "loss": 1.8598,
+      "step": 24040
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.8007556070320583e-05,
+      "loss": 1.8616,
+      "step": 24050
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.7986745787395333e-05,
+      "loss": 1.8413,
+      "step": 24060
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.796593647755804e-05,
+      "loss": 1.8168,
+      "step": 24070
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.794512815086673e-05,
+      "loss": 1.7945,
+      "step": 24080
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.792432081737887e-05,
+      "loss": 1.8828,
+      "step": 24090
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.7903514487151538e-05,
+      "loss": 1.8317,
+      "step": 24100
+    },
+    {
+      "epoch": 1.05,
+      "eval_loss": 1.8366378545761108,
+      "eval_runtime": 12.7159,
+      "eval_samples_per_second": 322.117,
+      "eval_steps_per_second": 20.132,
+      "step": 24100
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.7882709170241247e-05,
+      "loss": 1.8135,
+      "step": 24110
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.7861904876704087e-05,
+      "loss": 1.8198,
+      "step": 24120
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.7841101616595597e-05,
+      "loss": 1.8092,
+      "step": 24130
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.7820299399970864e-05,
+      "loss": 1.8312,
+      "step": 24140
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.779949823688443e-05,
+      "loss": 1.8219,
+      "step": 24150
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.777869813739036e-05,
+      "loss": 1.8105,
+      "step": 24160
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.775789911154218e-05,
+      "loss": 1.864,
+      "step": 24170
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.7737101169392924e-05,
+      "loss": 1.7589,
+      "step": 24180
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.7716304320995084e-05,
+      "loss": 1.8094,
+      "step": 24190
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.7695508576400596e-05,
+      "loss": 1.8204,
+      "step": 24200
+    },
+    {
+      "epoch": 1.06,
+      "eval_loss": 1.836849570274353,
+      "eval_runtime": 11.6165,
+      "eval_samples_per_second": 352.601,
+      "eval_steps_per_second": 22.038,
+      "step": 24200
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.767471394566094e-05,
+      "loss": 1.8652,
+      "step": 24210
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.7653920438826964e-05,
+      "loss": 1.8454,
+      "step": 24220
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.763312806594906e-05,
+      "loss": 1.7737,
+      "step": 24230
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.7612336837077e-05,
+      "loss": 1.7854,
+      "step": 24240
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.7591546762260058e-05,
+      "loss": 1.8266,
+      "step": 24250
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.7570757851546904e-05,
+      "loss": 1.7916,
+      "step": 24260
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.754997011498568e-05,
+      "loss": 1.8398,
+      "step": 24270
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.7529183562623943e-05,
+      "loss": 1.8321,
+      "step": 24280
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.7508398204508692e-05,
+      "loss": 1.8409,
+      "step": 24290
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.7487614050686325e-05,
+      "loss": 1.8149,
+      "step": 24300
+    },
+    {
+      "epoch": 1.06,
+      "eval_loss": 1.836502194404602,
+      "eval_runtime": 11.6567,
+      "eval_samples_per_second": 351.386,
+      "eval_steps_per_second": 21.962,
+      "step": 24300
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.746683111120269e-05,
+      "loss": 1.8265,
+      "step": 24310
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.744604939610301e-05,
+      "loss": 1.8582,
+      "step": 24320
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.7425268915431954e-05,
+      "loss": 1.8372,
+      "step": 24330
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.740448967923356e-05,
+      "loss": 1.8785,
+      "step": 24340
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.7383711697551307e-05,
+      "loss": 1.8081,
+      "step": 24350
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.7362934980428004e-05,
+      "loss": 1.8398,
+      "step": 24360
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.7342159537905934e-05,
+      "loss": 1.8358,
+      "step": 24370
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.7321385380026666e-05,
+      "loss": 1.8095,
+      "step": 24380
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.7300612516831248e-05,
+      "loss": 1.8128,
+      "step": 24390
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.727984095836001e-05,
+      "loss": 1.8152,
+      "step": 24400
+    },
+    {
+      "epoch": 1.07,
+      "eval_loss": 1.8360520601272583,
+      "eval_runtime": 13.1567,
+      "eval_samples_per_second": 311.323,
+      "eval_steps_per_second": 19.458,
+      "step": 24400
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.7259070714652725e-05,
+      "loss": 1.797,
+      "step": 24410
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.723830179574848e-05,
+      "loss": 1.8195,
+      "step": 24420
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.721753421168576e-05,
+      "loss": 1.8132,
+      "step": 24430
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.7196767972502376e-05,
+      "loss": 1.8053,
+      "step": 24440
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.717600308823549e-05,
+      "loss": 1.8321,
+      "step": 24450
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.7155239568921643e-05,
+      "loss": 1.8653,
+      "step": 24460
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.7134477424596657e-05,
+      "loss": 1.8642,
+      "step": 24470
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.711371666529577e-05,
+      "loss": 1.8297,
+      "step": 24480
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.7092957301053448e-05,
+      "loss": 1.8048,
+      "step": 24490
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.70721993419036e-05,
+      "loss": 1.8603,
+      "step": 24500
+    },
+    {
+      "epoch": 1.07,
+      "eval_loss": 1.8359553813934326,
+      "eval_runtime": 12.136,
+      "eval_samples_per_second": 337.508,
+      "eval_steps_per_second": 21.094,
+      "step": 24500
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.705144279787935e-05,
+      "loss": 1.8392,
+      "step": 24510
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.70306876790132e-05,
+      "loss": 1.8608,
+      "step": 24520
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.7009933995336948e-05,
+      "loss": 1.835,
+      "step": 24530
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.6989181756881696e-05,
+      "loss": 1.8186,
+      "step": 24540
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.6968430973677844e-05,
+      "loss": 1.8125,
+      "step": 24550
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.6947681655755097e-05,
+      "loss": 1.8156,
+      "step": 24560
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.6926933813142446e-05,
+      "loss": 1.824,
+      "step": 24570
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.690618745586817e-05,
+      "loss": 1.8335,
+      "step": 24580
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.6885442593959828e-05,
+      "loss": 1.7904,
+      "step": 24590
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.686469923744427e-05,
+      "loss": 1.8317,
+      "step": 24600
+    },
+    {
+      "epoch": 1.08,
+      "eval_loss": 1.8358409404754639,
+      "eval_runtime": 11.6144,
+      "eval_samples_per_second": 352.667,
+      "eval_steps_per_second": 22.042,
+      "step": 24600
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.6843957396347592e-05,
+      "loss": 1.7796,
+      "step": 24610
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.6823217080695197e-05,
+      "loss": 1.8364,
+      "step": 24620
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.6802478300511687e-05,
+      "loss": 1.8251,
+      "step": 24630
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.6781741065821003e-05,
+      "loss": 1.8535,
+      "step": 24640
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.676100538664626e-05,
+      "loss": 1.8141,
+      "step": 24650
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.6740271273009896e-05,
+      "loss": 1.8181,
+      "step": 24660
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.671953873493353e-05,
+      "loss": 1.8209,
+      "step": 24670
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.669880778243804e-05,
+      "loss": 1.7613,
+      "step": 24680
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.6678078425543563e-05,
+      "loss": 1.83,
+      "step": 24690
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.6657350674269424e-05,
+      "loss": 1.8316,
+      "step": 24700
+    },
+    {
+      "epoch": 1.08,
+      "eval_loss": 1.835756778717041,
+      "eval_runtime": 11.6376,
+      "eval_samples_per_second": 351.961,
+      "eval_steps_per_second": 21.998,
+      "step": 24700
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.6636624538634208e-05,
+      "loss": 1.8301,
+      "step": 24710
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.6615900028655687e-05,
+      "loss": 1.8184,
+      "step": 24720
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.6595177154350883e-05,
+      "loss": 1.8037,
+      "step": 24730
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.6574455925735977e-05,
+      "loss": 1.8436,
+      "step": 24740
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.6553736352826426e-05,
+      "loss": 1.8369,
+      "step": 24750
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.6533018445636797e-05,
+      "loss": 1.8328,
+      "step": 24760
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.651230221418095e-05,
+      "loss": 1.8164,
+      "step": 24770
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.6491587668471846e-05,
+      "loss": 1.8137,
+      "step": 24780
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.6470874818521696e-05,
+      "loss": 1.8131,
+      "step": 24790
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.6450163674341847e-05,
+      "loss": 1.8169,
+      "step": 24800
+    },
+    {
+      "epoch": 1.09,
+      "eval_loss": 1.8353500366210938,
+      "eval_runtime": 11.5666,
+      "eval_samples_per_second": 354.122,
+      "eval_steps_per_second": 22.133,
+      "step": 24800
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.6429454245942863e-05,
+      "loss": 1.8232,
+      "step": 24810
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.640874654333443e-05,
+      "loss": 1.7629,
+      "step": 24820
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.6388040576525452e-05,
+      "loss": 1.8425,
+      "step": 24830
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.636733635552395e-05,
+      "loss": 1.8757,
+      "step": 24840
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.634663389033713e-05,
+      "loss": 1.8382,
+      "step": 24850
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.632593319097133e-05,
+      "loss": 1.8147,
+      "step": 24860
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.6305234267432055e-05,
+      "loss": 1.8169,
+      "step": 24870
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.6284537129723926e-05,
+      "loss": 1.8293,
+      "step": 24880
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.6263841787850735e-05,
+      "loss": 1.8217,
+      "step": 24890
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.6243148251815364e-05,
+      "loss": 1.8505,
+      "step": 24900
+    },
+    {
+      "epoch": 1.09,
+      "eval_loss": 1.8351911306381226,
+      "eval_runtime": 13.3476,
+      "eval_samples_per_second": 306.872,
+      "eval_steps_per_second": 19.179,
+      "step": 24900
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.6222456531619848e-05,
+      "loss": 1.8278,
+      "step": 24910
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.6201766637265354e-05,
+      "loss": 1.8546,
+      "step": 24920
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.6181078578752133e-05,
+      "loss": 1.8576,
+      "step": 24930
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.6160392366079592e-05,
+      "loss": 1.8243,
+      "step": 24940
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.6139708009246196e-05,
+      "loss": 1.7829,
+      "step": 24950
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.611902551824956e-05,
+      "loss": 1.845,
+      "step": 24960
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.609834490308637e-05,
+      "loss": 1.8217,
+      "step": 24970
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.607766617375241e-05,
+      "loss": 1.8471,
+      "step": 24980
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.6056989340242555e-05,
+      "loss": 1.8651,
+      "step": 24990
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.6036314412550773e-05,
+      "loss": 1.795,
+      "step": 25000
+    },
+    {
+      "epoch": 1.09,
+      "eval_loss": 1.8348760604858398,
+      "eval_runtime": 11.8701,
+      "eval_samples_per_second": 345.068,
+      "eval_steps_per_second": 21.567,
+      "step": 25000
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.601564140067007e-05,
+      "loss": 1.7818,
+      "step": 25010
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.5994970314592603e-05,
+      "loss": 1.785,
+      "step": 25020
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.5974301164309514e-05,
+      "loss": 1.8312,
+      "step": 25030
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.595363395981106e-05,
+      "loss": 1.8278,
+      "step": 25040
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.593296871108654e-05,
+      "loss": 1.8119,
+      "step": 25050
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.591230542812433e-05,
+      "loss": 1.8369,
+      "step": 25060
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.5891644120911817e-05,
+      "loss": 1.8243,
+      "step": 25070
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.5870984799435472e-05,
+      "loss": 1.8056,
+      "step": 25080
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.585032747368077e-05,
+      "loss": 1.8389,
+      "step": 25090
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.582967215363226e-05,
+      "loss": 1.8413,
+      "step": 25100
+    },
+    {
+      "epoch": 1.1,
+      "eval_loss": 1.8349188566207886,
+      "eval_runtime": 12.9593,
+      "eval_samples_per_second": 316.066,
+      "eval_steps_per_second": 19.754,
+      "step": 25100
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.5809018849273484e-05,
+      "loss": 1.8708,
+      "step": 25110
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.5788367570587047e-05,
+      "loss": 1.8269,
+      "step": 25120
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.5767718327554532e-05,
+      "loss": 1.7818,
+      "step": 25130
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.5747071130156587e-05,
+      "loss": 1.8257,
+      "step": 25140
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.5726425988372825e-05,
+      "loss": 1.8555,
+      "step": 25150
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.5705782912181888e-05,
+      "loss": 1.8056,
+      "step": 25160
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.568514191156143e-05,
+      "loss": 1.8413,
+      "step": 25170
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.5664502996488072e-05,
+      "loss": 1.8206,
+      "step": 25180
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.5643866176937462e-05,
+      "loss": 1.868,
+      "step": 25190
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.5623231462884204e-05,
+      "loss": 1.8046,
+      "step": 25200
+    },
+    {
+      "epoch": 1.1,
+      "eval_loss": 1.8350028991699219,
+      "eval_runtime": 12.8788,
+      "eval_samples_per_second": 318.041,
+      "eval_steps_per_second": 19.878,
+      "step": 25200
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.5602598864301912e-05,
+      "loss": 1.8307,
+      "step": 25210
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.5581968391163145e-05,
+      "loss": 1.7947,
+      "step": 25220
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.5561340053439473e-05,
+      "loss": 1.8254,
+      "step": 25230
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.5540713861101395e-05,
+      "loss": 1.8283,
+      "step": 25240
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.5520089824118422e-05,
+      "loss": 1.8242,
+      "step": 25250
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.5499467952458946e-05,
+      "loss": 1.8352,
+      "step": 25260
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.547884825609041e-05,
+      "loss": 1.8657,
+      "step": 25270
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.545823074497911e-05,
+      "loss": 1.8535,
+      "step": 25280
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.5437615429090373e-05,
+      "loss": 1.8455,
+      "step": 25290
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.541700231838839e-05,
+      "loss": 1.8007,
+      "step": 25300
+    },
+    {
+      "epoch": 1.11,
+      "eval_loss": 1.8347127437591553,
+      "eval_runtime": 12.1617,
+      "eval_samples_per_second": 336.796,
+      "eval_steps_per_second": 21.05,
+      "step": 25300
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.5396391422836335e-05,
+      "loss": 1.8574,
+      "step": 25310
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.5375782752396286e-05,
+      "loss": 1.8717,
+      "step": 25320
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.535517631702927e-05,
+      "loss": 1.8575,
+      "step": 25330
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.5334572126695204e-05,
+      "loss": 1.8246,
+      "step": 25340
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.5313970191352945e-05,
+      "loss": 1.775,
+      "step": 25350
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.529337052096024e-05,
+      "loss": 1.834,
+      "step": 25360
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.527277312547376e-05,
+      "loss": 1.8294,
+      "step": 25370
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.5252178014849072e-05,
+      "loss": 1.785,
+      "step": 25380
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.523158519904061e-05,
+      "loss": 1.7699,
+      "step": 25390
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.521099468800175e-05,
+      "loss": 1.8178,
+      "step": 25400
+    },
+    {
+      "epoch": 1.11,
+      "eval_loss": 1.8346803188323975,
+      "eval_runtime": 11.7822,
+      "eval_samples_per_second": 347.642,
+      "eval_steps_per_second": 21.728,
+      "step": 25400
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.5190406491684708e-05,
+      "loss": 1.8216,
+      "step": 25410
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.5169820620040608e-05,
+      "loss": 1.8559,
+      "step": 25420
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.5149237083019438e-05,
+      "loss": 1.8424,
+      "step": 25430
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.5128655890570067e-05,
+      "loss": 1.8335,
+      "step": 25440
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.5108077052640223e-05,
+      "loss": 1.8141,
+      "step": 25450
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.5087500579176504e-05,
+      "loss": 1.851,
+      "step": 25460
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.506692648012435e-05,
+      "loss": 1.9212,
+      "step": 25470
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.5046354765428074e-05,
+      "loss": 1.845,
+      "step": 25480
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.502578544503081e-05,
+      "loss": 1.8246,
+      "step": 25490
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.5005218528874574e-05,
+      "loss": 1.8055,
+      "step": 25500
+    },
+    {
+      "epoch": 1.12,
+      "eval_loss": 1.834632158279419,
+      "eval_runtime": 11.8751,
+      "eval_samples_per_second": 344.925,
+      "eval_steps_per_second": 21.558,
+      "step": 25500
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.4984654026900173e-05,
+      "loss": 1.8475,
+      "step": 25510
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.4964091949047294e-05,
+      "loss": 1.8136,
+      "step": 25520
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.49435323052544e-05,
+      "loss": 1.7852,
+      "step": 25530
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.4922975105458838e-05,
+      "loss": 1.7962,
+      "step": 25540
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.4902420359596706e-05,
+      "loss": 1.8233,
+      "step": 25550
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.4881868077602996e-05,
+      "loss": 1.8193,
+      "step": 25560
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.4861318269411427e-05,
+      "loss": 1.8325,
+      "step": 25570
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.484077094495458e-05,
+      "loss": 1.8399,
+      "step": 25580
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.4820226114163803e-05,
+      "loss": 1.7936,
+      "step": 25590
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.4799683786969268e-05,
+      "loss": 1.7853,
+      "step": 25600
+    },
+    {
+      "epoch": 1.12,
+      "eval_loss": 1.834519863128662,
+      "eval_runtime": 11.6394,
+      "eval_samples_per_second": 351.907,
+      "eval_steps_per_second": 21.994,
+      "step": 25600
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.477914397329991e-05,
+      "loss": 1.8067,
+      "step": 25610
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.4758606683083456e-05,
+      "loss": 1.8196,
+      "step": 25620
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.4738071926246435e-05,
+      "loss": 1.7976,
+      "step": 25630
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.4717539712714103e-05,
+      "loss": 1.8813,
+      "step": 25640
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.4697010052410555e-05,
+      "loss": 1.7963,
+      "step": 25650
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.4676482955258578e-05,
+      "loss": 1.8388,
+      "step": 25660
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.465595843117979e-05,
+      "loss": 1.8194,
+      "step": 25670
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.46354364900945e-05,
+      "loss": 1.83,
+      "step": 25680
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.461491714192182e-05,
+      "loss": 1.8395,
+      "step": 25690
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.4594400396579582e-05,
+      "loss": 1.8314,
+      "step": 25700
+    },
+    {
+      "epoch": 1.13,
+      "eval_loss": 1.8341410160064697,
+      "eval_runtime": 13.3901,
+      "eval_samples_per_second": 305.899,
+      "eval_steps_per_second": 19.119,
+      "step": 25700
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.457388626398437e-05,
+      "loss": 1.8478,
+      "step": 25710
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.4553374754051494e-05,
+      "loss": 1.8013,
+      "step": 25720
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.4532865876695022e-05,
+      "loss": 1.8231,
+      "step": 25730
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.4512359641827707e-05,
+      "loss": 1.8409,
+      "step": 25740
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.4491856059361073e-05,
+      "loss": 1.8713,
+      "step": 25750
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.4471355139205323e-05,
+      "loss": 1.8156,
+      "step": 25760
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.44508568912694e-05,
+      "loss": 1.8367,
+      "step": 25770
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.4430361325460937e-05,
+      "loss": 1.8107,
+      "step": 25780
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.440986845168629e-05,
+      "loss": 1.7991,
+      "step": 25790
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.438937827985047e-05,
+      "loss": 1.8464,
+      "step": 25800
+    },
+    {
+      "epoch": 1.13,
+      "eval_loss": 1.834221601486206,
+      "eval_runtime": 11.7029,
+      "eval_samples_per_second": 349.999,
+      "eval_steps_per_second": 21.875,
+      "step": 25800
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.4368890819857256e-05,
+      "loss": 1.8099,
+      "step": 25810
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.4348406081609035e-05,
+      "loss": 1.8311,
+      "step": 25820
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.432792407500695e-05,
+      "loss": 1.8057,
+      "step": 25830
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.4307444809950767e-05,
+      "loss": 1.8371,
+      "step": 25840
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.428696829633897e-05,
+      "loss": 1.8405,
+      "step": 25850
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.4266494544068682e-05,
+      "loss": 1.8629,
+      "step": 25860
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.4246023563035703e-05,
+      "loss": 1.8636,
+      "step": 25870
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.4225555363134504e-05,
+      "loss": 1.819,
+      "step": 25880
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.4205089954258187e-05,
+      "loss": 1.8154,
+      "step": 25890
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.4184627346298544e-05,
+      "loss": 1.8101,
+      "step": 25900
+    },
+    {
+      "epoch": 1.13,
+      "eval_loss": 1.8341854810714722,
+      "eval_runtime": 11.5933,
+      "eval_samples_per_second": 353.308,
+      "eval_steps_per_second": 22.082,
+      "step": 25900
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.4164167549145952e-05,
+      "loss": 1.8222,
+      "step": 25910
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.4143710572689507e-05,
+      "loss": 1.8402,
+      "step": 25920
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.4123256426816873e-05,
+      "loss": 1.8229,
+      "step": 25930
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.4102805121414386e-05,
+      "loss": 1.7893,
+      "step": 25940
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.4082356666366987e-05,
+      "loss": 1.8193,
+      "step": 25950
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.4061911071558266e-05,
+      "loss": 1.8274,
+      "step": 25960
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.4041468346870393e-05,
+      "loss": 1.8479,
+      "step": 25970
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.4021028502184194e-05,
+      "loss": 1.8438,
+      "step": 25980
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.4000591547379058e-05,
+      "loss": 1.8017,
+      "step": 25990
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.398015749233302e-05,
+      "loss": 1.8014,
+      "step": 26000
+    },
+    {
+      "epoch": 1.14,
+      "eval_loss": 1.8340920209884644,
+      "eval_runtime": 11.695,
+      "eval_samples_per_second": 350.235,
+      "eval_steps_per_second": 21.89,
+      "step": 26000
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.3959726346922674e-05,
+      "loss": 1.7826,
+      "step": 26010
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.393929812102325e-05,
+      "loss": 1.8292,
+      "step": 26020
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.391887282450852e-05,
+      "loss": 1.8536,
+      "step": 26030
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.389845046725089e-05,
+      "loss": 1.8431,
+      "step": 26040
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.3878031059121282e-05,
+      "loss": 1.8513,
+      "step": 26050
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.3857614609989278e-05,
+      "loss": 1.8321,
+      "step": 26060
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.383720112972293e-05,
+      "loss": 1.8147,
+      "step": 26070
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.3816790628188953e-05,
+      "loss": 1.7952,
+      "step": 26080
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.3796383115252546e-05,
+      "loss": 1.8118,
+      "step": 26090
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.37759786007775e-05,
+      "loss": 1.8081,
+      "step": 26100
+    },
+    {
+      "epoch": 1.14,
+      "eval_loss": 1.833876371383667,
+      "eval_runtime": 11.5373,
+      "eval_samples_per_second": 355.021,
+      "eval_steps_per_second": 22.189,
+      "step": 26100
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.3755577094626154e-05,
+      "loss": 1.8162,
+      "step": 26110
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.3735178606659382e-05,
+      "loss": 1.8302,
+      "step": 26120
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.3714783146736622e-05,
+      "loss": 1.8093,
+      "step": 26130
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.369439072471581e-05,
+      "loss": 1.8492,
+      "step": 26140
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.3674001350453468e-05,
+      "loss": 1.8483,
+      "step": 26150
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.3653615033804564e-05,
+      "loss": 1.8476,
+      "step": 26160
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.3633231784622693e-05,
+      "loss": 1.8372,
+      "step": 26170
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.3612851612759865e-05,
+      "loss": 1.8233,
+      "step": 26180
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.3592474528066693e-05,
+      "loss": 1.7858,
+      "step": 26190
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.3572100540392215e-05,
+      "loss": 1.8376,
+      "step": 26200
+    },
+    {
+      "epoch": 1.15,
+      "eval_loss": 1.833698034286499,
+      "eval_runtime": 11.6849,
+      "eval_samples_per_second": 350.538,
+      "eval_steps_per_second": 21.909,
+      "step": 26200
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.355172965958403e-05,
+      "loss": 1.8245,
+      "step": 26210
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.353136189548821e-05,
+      "loss": 1.842,
+      "step": 26220
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.3510997257949336e-05,
+      "loss": 1.8888,
+      "step": 26230
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.3490635756810458e-05,
+      "loss": 1.8482,
+      "step": 26240
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.3470277401913134e-05,
+      "loss": 1.8517,
+      "step": 26250
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.3449922203097373e-05,
+      "loss": 1.8199,
+      "step": 26260
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.342957017020169e-05,
+      "loss": 1.824,
+      "step": 26270
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.3409221313063035e-05,
+      "loss": 1.8576,
+      "step": 26280
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.3388875641516863e-05,
+      "loss": 1.8216,
+      "step": 26290
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.3368533165397046e-05,
+      "loss": 1.8346,
+      "step": 26300
+    },
+    {
+      "epoch": 1.15,
+      "eval_loss": 1.8334472179412842,
+      "eval_runtime": 11.5788,
+      "eval_samples_per_second": 353.751,
+      "eval_steps_per_second": 22.109,
+      "step": 26300
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.3348193894535956e-05,
+      "loss": 1.7762,
+      "step": 26310
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.332785783876438e-05,
+      "loss": 1.8077,
+      "step": 26320
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.3307525007911557e-05,
+      "loss": 1.8083,
+      "step": 26330
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.328719541180519e-05,
+      "loss": 1.8251,
+      "step": 26340
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.3266869060271383e-05,
+      "loss": 1.8575,
+      "step": 26350
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.3246545963134718e-05,
+      "loss": 1.7965,
+      "step": 26360
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.3226226130218155e-05,
+      "loss": 1.8171,
+      "step": 26370
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.3205909571343112e-05,
+      "loss": 1.7914,
+      "step": 26380
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.3185596296329404e-05,
+      "loss": 1.8391,
+      "step": 26390
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.3165286314995278e-05,
+      "loss": 1.7914,
+      "step": 26400
+    },
+    {
+      "epoch": 1.16,
+      "eval_loss": 1.8334349393844604,
+      "eval_runtime": 11.5307,
+      "eval_samples_per_second": 355.226,
+      "eval_steps_per_second": 22.202,
+      "step": 26400
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.3144979637157365e-05,
+      "loss": 1.8035,
+      "step": 26410
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.3124676272630724e-05,
+      "loss": 1.8065,
+      "step": 26420
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.3104376231228773e-05,
+      "loss": 1.7798,
+      "step": 26430
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.308407952276339e-05,
+      "loss": 1.7659,
+      "step": 26440
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.3063786157044766e-05,
+      "loss": 1.8257,
+      "step": 26450
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.3043496143881546e-05,
+      "loss": 1.8107,
+      "step": 26460
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.3023209493080696e-05,
+      "loss": 1.8759,
+      "step": 26470
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.3002926214447598e-05,
+      "loss": 1.8089,
+      "step": 26480
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.2982646317785977e-05,
+      "loss": 1.7965,
+      "step": 26490
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.296236981289795e-05,
+      "loss": 1.8116,
+      "step": 26500
+    },
+    {
+      "epoch": 1.16,
+      "eval_loss": 1.8330750465393066,
+      "eval_runtime": 12.1921,
+      "eval_samples_per_second": 335.955,
+      "eval_steps_per_second": 20.997,
+      "step": 26500
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.2942096709583966e-05,
+      "loss": 1.8146,
+      "step": 26510
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.2921827017642855e-05,
+      "loss": 1.8026,
+      "step": 26520
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.2901560746871777e-05,
+      "loss": 1.8253,
+      "step": 26530
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.2881297907066254e-05,
+      "loss": 1.786,
+      "step": 26540
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.286103850802014e-05,
+      "loss": 1.8141,
+      "step": 26550
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.2840782559525635e-05,
+      "loss": 1.8312,
+      "step": 26560
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.2820530071373278e-05,
+      "loss": 1.7801,
+      "step": 26570
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.280028105335189e-05,
+      "loss": 1.8628,
+      "step": 26580
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.2780035515248673e-05,
+      "loss": 1.8254,
+      "step": 26590
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.275979346684911e-05,
+      "loss": 1.8014,
+      "step": 26600
+    },
+    {
+      "epoch": 1.16,
+      "eval_loss": 1.832726001739502,
+      "eval_runtime": 11.4701,
+      "eval_samples_per_second": 357.102,
+      "eval_steps_per_second": 22.319,
+      "step": 26600
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.273955491793702e-05,
+      "loss": 1.8464,
+      "step": 26610
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.2719319878294514e-05,
+      "loss": 1.854,
+      "step": 26620
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.2699088357702025e-05,
+      "loss": 1.8389,
+      "step": 26630
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.267886036593825e-05,
+      "loss": 1.7847,
+      "step": 26640
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.2658635912780227e-05,
+      "loss": 1.8347,
+      "step": 26650
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.2638415008003243e-05,
+      "loss": 1.7616,
+      "step": 26660
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.2618197661380898e-05,
+      "loss": 1.7771,
+      "step": 26670
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.259798388268505e-05,
+      "loss": 1.8119,
+      "step": 26680
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.2577773681685867e-05,
+      "loss": 1.8511,
+      "step": 26690
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.2557567068151724e-05,
+      "loss": 1.7993,
+      "step": 26700
+    },
+    {
+      "epoch": 1.17,
+      "eval_loss": 1.8325154781341553,
+      "eval_runtime": 11.4069,
+      "eval_samples_per_second": 359.08,
+      "eval_steps_per_second": 22.443,
+      "step": 26700
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.253736405184935e-05,
+      "loss": 1.7621,
+      "step": 26710
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.2517164642543642e-05,
+      "loss": 1.8126,
+      "step": 26720
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.249696884999784e-05,
+      "loss": 1.826,
+      "step": 26730
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.2476776683973364e-05,
+      "loss": 1.7861,
+      "step": 26740
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.2456588154229932e-05,
+      "loss": 1.8022,
+      "step": 26750
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.2436403270525473e-05,
+      "loss": 1.7953,
+      "step": 26760
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.2416222042616177e-05,
+      "loss": 1.8091,
+      "step": 26770
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.2396044480256443e-05,
+      "loss": 1.839,
+      "step": 26780
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.237587059319892e-05,
+      "loss": 1.835,
+      "step": 26790
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.235570039119448e-05,
+      "loss": 1.8417,
+      "step": 26800
+    },
+    {
+      "epoch": 1.17,
+      "eval_loss": 1.8325867652893066,
+      "eval_runtime": 11.4762,
+      "eval_samples_per_second": 356.911,
+      "eval_steps_per_second": 22.307,
+      "step": 26800
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.2335533883992166e-05,
+      "loss": 1.8167,
+      "step": 26810
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.2315371081339328e-05,
+      "loss": 1.8286,
+      "step": 26820
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.2295211992981426e-05,
+      "loss": 1.862,
+      "step": 26830
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.2275056628662205e-05,
+      "loss": 1.7991,
+      "step": 26840
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.225490499812355e-05,
+      "loss": 1.8163,
+      "step": 26850
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.2234757111105584e-05,
+      "loss": 1.8737,
+      "step": 26860
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.2214612977346593e-05,
+      "loss": 1.8203,
+      "step": 26870
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.2194472606583074e-05,
+      "loss": 1.8228,
+      "step": 26880
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.2174336008549667e-05,
+      "loss": 1.8561,
+      "step": 26890
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.2154203192979235e-05,
+      "loss": 1.8229,
+      "step": 26900
+    },
+    {
+      "epoch": 1.18,
+      "eval_loss": 1.8322032690048218,
+      "eval_runtime": 11.5781,
+      "eval_samples_per_second": 353.771,
+      "eval_steps_per_second": 22.111,
+      "step": 26900
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.2134074169602773e-05,
+      "loss": 1.8383,
+      "step": 26910
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.2113948948149477e-05,
+      "loss": 1.792,
+      "step": 26920
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.209382753834667e-05,
+      "loss": 1.7557,
+      "step": 26930
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.2073709949919867e-05,
+      "loss": 1.8174,
+      "step": 26940
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.2053596192592704e-05,
+      "loss": 1.8101,
+      "step": 26950
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.2033486276087e-05,
+      "loss": 1.8121,
+      "step": 26960
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.201338021012268e-05,
+      "loss": 1.792,
+      "step": 26970
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.199327800441785e-05,
+      "loss": 1.8288,
+      "step": 26980
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.1973179668688697e-05,
+      "loss": 1.8468,
+      "step": 26990
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.195308521264959e-05,
+      "loss": 1.8541,
+      "step": 27000
+    },
+    {
+      "epoch": 1.18,
+      "eval_loss": 1.8320529460906982,
+      "eval_runtime": 11.853,
+      "eval_samples_per_second": 345.568,
+      "eval_steps_per_second": 21.598,
+      "step": 27000
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.193299464601299e-05,
+      "loss": 1.8016,
+      "step": 27010
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.1912907978489493e-05,
+      "loss": 1.8023,
+      "step": 27020
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.1892825219787804e-05,
+      "loss": 1.8119,
+      "step": 27030
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.1872746379614736e-05,
+      "loss": 1.808,
+      "step": 27040
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.185267146767522e-05,
+      "loss": 1.8235,
+      "step": 27050
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.1832600493672274e-05,
+      "loss": 1.8342,
+      "step": 27060
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.1812533467307035e-05,
+      "loss": 1.8602,
+      "step": 27070
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.1792470398278683e-05,
+      "loss": 1.8775,
+      "step": 27080
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.177241129628456e-05,
+      "loss": 1.8305,
+      "step": 27090
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.1752356171020016e-05,
+      "loss": 1.8182,
+      "step": 27100
+    },
+    {
+      "epoch": 1.19,
+      "eval_loss": 1.832180142402649,
+      "eval_runtime": 11.4296,
+      "eval_samples_per_second": 358.368,
+      "eval_steps_per_second": 22.398,
+      "step": 27100
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.1732305032178533e-05,
+      "loss": 1.8361,
+      "step": 27110
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.1712257889451627e-05,
+      "loss": 1.849,
+      "step": 27120
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.1692214752528916e-05,
+      "loss": 1.8339,
+      "step": 27130
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.1672175631098056e-05,
+      "loss": 1.8372,
+      "step": 27140
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.1652140534844787e-05,
+      "loss": 1.7972,
+      "step": 27150
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.1632109473452864e-05,
+      "loss": 1.8092,
+      "step": 27160
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.161208245660415e-05,
+      "loss": 1.8685,
+      "step": 27170
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.1592059493978492e-05,
+      "loss": 1.8396,
+      "step": 27180
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.1572040595253822e-05,
+      "loss": 1.7891,
+      "step": 27190
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.1552025770106077e-05,
+      "loss": 1.7929,
+      "step": 27200
+    },
+    {
+      "epoch": 1.19,
+      "eval_loss": 1.8319555521011353,
+      "eval_runtime": 11.6871,
+      "eval_samples_per_second": 350.47,
+      "eval_steps_per_second": 21.904,
+      "step": 27200
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.1532015028209264e-05,
+      "loss": 1.8318,
+      "step": 27210
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.1512008379235355e-05,
+      "loss": 1.8183,
+      "step": 27220
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.149200583285442e-05,
+      "loss": 1.8565,
+      "step": 27230
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.1472007398734464e-05,
+      "loss": 1.8268,
+      "step": 27240
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.1452013086541593e-05,
+      "loss": 1.7994,
+      "step": 27250
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.143202290593984e-05,
+      "loss": 1.8272,
+      "step": 27260
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.1412036866591293e-05,
+      "loss": 1.8074,
+      "step": 27270
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.1392054978156015e-05,
+      "loss": 1.8095,
+      "step": 27280
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.137207725029206e-05,
+      "loss": 1.8097,
+      "step": 27290
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.1352103692655497e-05,
+      "loss": 1.7954,
+      "step": 27300
+    },
+    {
+      "epoch": 1.2,
+      "eval_loss": 1.8318532705307007,
+      "eval_runtime": 11.5752,
+      "eval_samples_per_second": 353.86,
+      "eval_steps_per_second": 22.116,
+      "step": 27300
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.133213431490035e-05,
+      "loss": 1.809,
+      "step": 27310
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.1312169126678647e-05,
+      "loss": 1.8336,
+      "step": 27320
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.129220813764035e-05,
+      "loss": 1.8235,
+      "step": 27330
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.127225135743346e-05,
+      "loss": 1.7739,
+      "step": 27340
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.1252298795703863e-05,
+      "loss": 1.857,
+      "step": 27350
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.123235046209549e-05,
+      "loss": 1.7995,
+      "step": 27360
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.121240636625015e-05,
+      "loss": 1.8557,
+      "step": 27370
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.1192466517807657e-05,
+      "loss": 1.8035,
+      "step": 27380
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.1172530926405745e-05,
+      "loss": 1.8037,
+      "step": 27390
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.1152599601680105e-05,
+      "loss": 1.8958,
+      "step": 27400
+    },
+    {
+      "epoch": 1.2,
+      "eval_loss": 1.8316702842712402,
+      "eval_runtime": 11.651,
+      "eval_samples_per_second": 351.557,
+      "eval_steps_per_second": 21.972,
+      "step": 27400
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.1132672553264356e-05,
+      "loss": 1.8168,
+      "step": 27410
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.111274979079006e-05,
+      "loss": 1.8574,
+      "step": 27420
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.1092831323886694e-05,
+      "loss": 1.7885,
+      "step": 27430
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.1072917162181672e-05,
+      "loss": 1.8138,
+      "step": 27440
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.105300731530032e-05,
+      "loss": 1.8419,
+      "step": 27450
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.1033101792865885e-05,
+      "loss": 1.7851,
+      "step": 27460
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.1013200604499507e-05,
+      "loss": 1.8412,
+      "step": 27470
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.099330375982026e-05,
+      "loss": 1.8451,
+      "step": 27480
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.0973411268445075e-05,
+      "loss": 1.8418,
+      "step": 27490
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.095352313998884e-05,
+      "loss": 1.8237,
+      "step": 27500
+    },
+    {
+      "epoch": 1.2,
+      "eval_loss": 1.831519603729248,
+      "eval_runtime": 11.976,
+      "eval_samples_per_second": 342.019,
+      "eval_steps_per_second": 21.376,
+      "step": 27500
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.0933639384064277e-05,
+      "loss": 1.8165,
+      "step": 27510
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.0913760010282015e-05,
+      "loss": 1.8361,
+      "step": 27520
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.0893885028250577e-05,
+      "loss": 1.8167,
+      "step": 27530
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.0874014447576348e-05,
+      "loss": 1.7941,
+      "step": 27540
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.0854148277863597e-05,
+      "loss": 1.799,
+      "step": 27550
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.0834286528714445e-05,
+      "loss": 1.8005,
+      "step": 27560
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.08144292097289e-05,
+      "loss": 1.8535,
+      "step": 27570
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.0794576330504802e-05,
+      "loss": 1.8461,
+      "step": 27580
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.0774727900637875e-05,
+      "loss": 1.8137,
+      "step": 27590
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.0754883929721646e-05,
+      "loss": 1.8526,
+      "step": 27600
+    },
+    {
+      "epoch": 1.21,
+      "eval_loss": 1.8314945697784424,
+      "eval_runtime": 11.6983,
+      "eval_samples_per_second": 350.137,
+      "eval_steps_per_second": 21.884,
+      "step": 27600
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.0735044427347557e-05,
+      "loss": 1.7918,
+      "step": 27610
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.0715209403104805e-05,
+      "loss": 1.8192,
+      "step": 27620
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.0695378866580508e-05,
+      "loss": 1.8595,
+      "step": 27630
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.0675552827359544e-05,
+      "loss": 1.8632,
+      "step": 27640
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.065573129502467e-05,
+      "loss": 1.8333,
+      "step": 27650
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.0635914279156423e-05,
+      "loss": 1.7787,
+      "step": 27660
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.0616101789333192e-05,
+      "loss": 1.7813,
+      "step": 27670
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.0596293835131144e-05,
+      "loss": 1.7958,
+      "step": 27680
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.057649042612429e-05,
+      "loss": 1.8267,
+      "step": 27690
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.0556691571884413e-05,
+      "loss": 1.808,
+      "step": 27700
+    },
+    {
+      "epoch": 1.21,
+      "eval_loss": 1.8313714265823364,
+      "eval_runtime": 11.5536,
+      "eval_samples_per_second": 354.521,
+      "eval_steps_per_second": 22.158,
+      "step": 27700
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.0536897281981125e-05,
+      "loss": 1.8536,
+      "step": 27710
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.0517107565981794e-05,
+      "loss": 1.8201,
+      "step": 27720
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.0497322433451612e-05,
+      "loss": 1.8408,
+      "step": 27730
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.0477541893953545e-05,
+      "loss": 1.8357,
+      "step": 27740
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.0457765957048314e-05,
+      "loss": 1.7831,
+      "step": 27750
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.0437994632294456e-05,
+      "loss": 1.788,
+      "step": 27760
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.0418227929248246e-05,
+      "loss": 1.833,
+      "step": 27770
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.0398465857463757e-05,
+      "loss": 1.8343,
+      "step": 27780
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.0378708426492782e-05,
+      "loss": 1.8193,
+      "step": 27790
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.035895564588492e-05,
+      "loss": 1.8208,
+      "step": 27800
+    },
+    {
+      "epoch": 1.22,
+      "eval_loss": 1.8313522338867188,
+      "eval_runtime": 11.4654,
+      "eval_samples_per_second": 357.248,
+      "eval_steps_per_second": 22.328,
+      "step": 27800
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.0339207525187474e-05,
+      "loss": 1.8266,
+      "step": 27810
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.031946407394553e-05,
+      "loss": 1.7958,
+      "step": 27820
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.02997253017019e-05,
+      "loss": 1.8288,
+      "step": 27830
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.027999121799714e-05,
+      "loss": 1.8183,
+      "step": 27840
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.0260261832369536e-05,
+      "loss": 1.8376,
+      "step": 27850
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.0240537154355117e-05,
+      "loss": 1.8534,
+      "step": 27860
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.0220817193487602e-05,
+      "loss": 1.827,
+      "step": 27870
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.0201101959298482e-05,
+      "loss": 1.8304,
+      "step": 27880
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.018139146131691e-05,
+      "loss": 1.8339,
+      "step": 27890
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.0161685709069786e-05,
+      "loss": 1.7982,
+      "step": 27900
+    },
+    {
+      "epoch": 1.22,
+      "eval_loss": 1.8311944007873535,
+      "eval_runtime": 11.5326,
+      "eval_samples_per_second": 355.168,
+      "eval_steps_per_second": 22.198,
+      "step": 27900
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.0141984712081704e-05,
+      "loss": 1.8171,
+      "step": 27910
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.012228847987496e-05,
+      "loss": 1.8211,
+      "step": 27920
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.010259702196954e-05,
+      "loss": 1.8328,
+      "step": 27930
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.0082910347883144e-05,
+      "loss": 1.7907,
+      "step": 27940
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.0063228467131136e-05,
+      "loss": 1.8231,
+      "step": 27950
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.0043551389226576e-05,
+      "loss": 1.8691,
+      "step": 27960
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.0023879123680192e-05,
+      "loss": 1.8043,
+      "step": 27970
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.000421168000041e-05,
+      "loss": 1.8543,
+      "step": 27980
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.9984549067693303e-05,
+      "loss": 1.8065,
+      "step": 27990
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.9964891296262595e-05,
+      "loss": 1.8391,
+      "step": 28000
+    },
+    {
+      "epoch": 1.23,
+      "eval_loss": 1.8308237791061401,
+      "eval_runtime": 11.5213,
+      "eval_samples_per_second": 355.515,
+      "eval_steps_per_second": 22.22,
+      "step": 28000
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.9945238375209708e-05,
+      "loss": 1.8262,
+      "step": 28010
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.992559031403369e-05,
+      "loss": 1.8074,
+      "step": 28020
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.9905947122231273e-05,
+      "loss": 1.792,
+      "step": 28030
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.988630880929679e-05,
+      "loss": 1.7933,
+      "step": 28040
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.9866675384722252e-05,
+      "loss": 1.8644,
+      "step": 28050
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.9847046857997285e-05,
+      "loss": 1.8546,
+      "step": 28060
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.9827423238609173e-05,
+      "loss": 1.8298,
+      "step": 28070
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.9807804536042796e-05,
+      "loss": 1.8519,
+      "step": 28080
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.978819075978069e-05,
+      "loss": 1.8168,
+      "step": 28090
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.9768581919302978e-05,
+      "loss": 1.8294,
+      "step": 28100
+    },
+    {
+      "epoch": 1.23,
+      "eval_loss": 1.830810785293579,
+      "eval_runtime": 11.61,
+      "eval_samples_per_second": 352.798,
+      "eval_steps_per_second": 22.05,
+      "step": 28100
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.9748978024087435e-05,
+      "loss": 1.8529,
+      "step": 28110
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.972937908360939e-05,
+      "loss": 1.8372,
+      "step": 28120
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.970978510734185e-05,
+      "loss": 1.8326,
+      "step": 28130
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.969019610475535e-05,
+      "loss": 1.8487,
+      "step": 28140
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.9670612085318082e-05,
+      "loss": 1.8219,
+      "step": 28150
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.9651033058495783e-05,
+      "loss": 1.837,
+      "step": 28160
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.963145903375181e-05,
+      "loss": 1.8009,
+      "step": 28170
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.9611890020547075e-05,
+      "loss": 1.8739,
+      "step": 28180
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.9592326028340093e-05,
+      "loss": 1.8071,
+      "step": 28190
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.9572767066586933e-05,
+      "loss": 1.7504,
+      "step": 28200
+    },
+    {
+      "epoch": 1.23,
+      "eval_loss": 1.830684781074524,
+      "eval_runtime": 11.5059,
+      "eval_samples_per_second": 355.991,
+      "eval_steps_per_second": 22.249,
+      "step": 28200
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.955321314474124e-05,
+      "loss": 1.8246,
+      "step": 28210
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.9533664272254245e-05,
+      "loss": 1.8368,
+      "step": 28220
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.9514120458574667e-05,
+      "loss": 1.8114,
+      "step": 28230
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.9494581713148883e-05,
+      "loss": 1.7932,
+      "step": 28240
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.9475048045420723e-05,
+      "loss": 1.8443,
+      "step": 28250
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.9455519464831643e-05,
+      "loss": 1.8332,
+      "step": 28260
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.9435995980820576e-05,
+      "loss": 1.8698,
+      "step": 28270
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.9416477602824035e-05,
+      "loss": 1.8356,
+      "step": 28280
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.9396964340276034e-05,
+      "loss": 1.8133,
+      "step": 28290
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.9377456202608148e-05,
+      "loss": 1.8127,
+      "step": 28300
+    },
+    {
+      "epoch": 1.24,
+      "eval_loss": 1.830805778503418,
+      "eval_runtime": 11.4938,
+      "eval_samples_per_second": 356.366,
+      "eval_steps_per_second": 22.273,
+      "step": 28300
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.935795319924944e-05,
+      "loss": 1.7979,
+      "step": 28310
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.933845533962652e-05,
+      "loss": 1.8191,
+      "step": 28320
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.9318962633163493e-05,
+      "loss": 1.8228,
+      "step": 28330
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.9299475089281988e-05,
+      "loss": 1.8554,
+      "step": 28340
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.9279992717401114e-05,
+      "loss": 1.8253,
+      "step": 28350
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.9260515526937517e-05,
+      "loss": 1.8192,
+      "step": 28360
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.9241043527305306e-05,
+      "loss": 1.8501,
+      "step": 28370
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.9221576727916107e-05,
+      "loss": 1.8767,
+      "step": 28380
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.920211513817899e-05,
+      "loss": 1.8047,
+      "step": 28390
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.918265876750057e-05,
+      "loss": 1.8454,
+      "step": 28400
+    },
+    {
+      "epoch": 1.24,
+      "eval_loss": 1.8304309844970703,
+      "eval_runtime": 11.674,
+      "eval_samples_per_second": 350.864,
+      "eval_steps_per_second": 21.929,
+      "step": 28400
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.9163207625284877e-05,
+      "loss": 1.8197,
+      "step": 28410
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.9143761720933473e-05,
+      "loss": 1.849,
+      "step": 28420
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.9124321063845325e-05,
+      "loss": 1.8015,
+      "step": 28430
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.910488566341692e-05,
+      "loss": 1.8752,
+      "step": 28440
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.9085455529042175e-05,
+      "loss": 1.834,
+      "step": 28450
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.9066030670112456e-05,
+      "loss": 1.8128,
+      "step": 28460
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.9046611096016604e-05,
+      "loss": 1.8009,
+      "step": 28470
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.9027196816140885e-05,
+      "loss": 1.7636,
+      "step": 28480
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.9007787839869026e-05,
+      "loss": 1.8176,
+      "step": 28490
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.8988384176582147e-05,
+      "loss": 1.8369,
+      "step": 28500
+    },
+    {
+      "epoch": 1.25,
+      "eval_loss": 1.8305020332336426,
+      "eval_runtime": 11.6462,
+      "eval_samples_per_second": 351.703,
+      "eval_steps_per_second": 21.981,
+      "step": 28500
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.8968985835658877e-05,
+      "loss": 1.7947,
+      "step": 28510
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.8949592826475173e-05,
+      "loss": 1.7933,
+      "step": 28520
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.893020515840452e-05,
+      "loss": 1.8181,
+      "step": 28530
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.8910822840817733e-05,
+      "loss": 1.8049,
+      "step": 28540
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.8891445883083093e-05,
+      "loss": 1.8141,
+      "step": 28550
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.8872074294566265e-05,
+      "loss": 1.8061,
+      "step": 28560
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.8852708084630347e-05,
+      "loss": 1.8378,
+      "step": 28570
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.88333472626358e-05,
+      "loss": 1.8434,
+      "step": 28580
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.881399183794051e-05,
+      "loss": 1.7783,
+      "step": 28590
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.879464181989974e-05,
+      "loss": 1.8562,
+      "step": 28600
+    },
+    {
+      "epoch": 1.25,
+      "eval_loss": 1.8305237293243408,
+      "eval_runtime": 11.6453,
+      "eval_samples_per_second": 351.728,
+      "eval_steps_per_second": 21.983,
+      "step": 28600
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.877529721786615e-05,
+      "loss": 1.8491,
+      "step": 28610
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.8755958041189765e-05,
+      "loss": 1.835,
+      "step": 28620
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.8736624299218016e-05,
+      "loss": 1.7899,
+      "step": 28630
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.8717296001295673e-05,
+      "loss": 1.7965,
+      "step": 28640
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.8697973156764906e-05,
+      "loss": 1.7844,
+      "step": 28650
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.867865577496521e-05,
+      "loss": 1.8755,
+      "step": 28660
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.8659343865233497e-05,
+      "loss": 1.7975,
+      "step": 28670
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.8640037436903977e-05,
+      "loss": 1.8462,
+      "step": 28680
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.8620736499308255e-05,
+      "loss": 1.833,
+      "step": 28690
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.8601441061775248e-05,
+      "loss": 1.8273,
+      "step": 28700
+    },
+    {
+      "epoch": 1.26,
+      "eval_loss": 1.830354928970337,
+      "eval_runtime": 11.5397,
+      "eval_samples_per_second": 354.948,
+      "eval_steps_per_second": 22.184,
+      "step": 28700
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.8582151133631224e-05,
+      "loss": 1.7919,
+      "step": 28710
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.856286672419981e-05,
+      "loss": 1.8614,
+      "step": 28720
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.8543587842801934e-05,
+      "loss": 1.8111,
+      "step": 28730
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.8524314498755872e-05,
+      "loss": 1.8388,
+      "step": 28740
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.8505046701377218e-05,
+      "loss": 1.7883,
+      "step": 28750
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.84857844599789e-05,
+      "loss": 1.8307,
+      "step": 28760
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.846652778387111e-05,
+      "loss": 1.8396,
+      "step": 28770
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.8447276682361428e-05,
+      "loss": 1.8378,
+      "step": 28780
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.842803116475466e-05,
+      "loss": 1.8442,
+      "step": 28790
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.8408791240352983e-05,
+      "loss": 1.7593,
+      "step": 28800
+    },
+    {
+      "epoch": 1.26,
+      "eval_loss": 1.8302619457244873,
+      "eval_runtime": 11.3952,
+      "eval_samples_per_second": 359.45,
+      "eval_steps_per_second": 22.466,
+      "step": 28800
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.8389556918455813e-05,
+      "loss": 1.8163,
+      "step": 28810
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.83703282083599e-05,
+      "loss": 1.7921,
+      "step": 28820
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.835110511935925e-05,
+      "loss": 1.8257,
+      "step": 28830
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.8331887660745182e-05,
+      "loss": 1.7806,
+      "step": 28840
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.8312675841806263e-05,
+      "loss": 1.8569,
+      "step": 28850
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.8293469671828365e-05,
+      "loss": 1.8384,
+      "step": 28860
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.82742691600946e-05,
+      "loss": 1.8023,
+      "step": 28870
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.8255074315885372e-05,
+      "loss": 1.8437,
+      "step": 28880
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.8235885148478325e-05,
+      "loss": 1.7939,
+      "step": 28890
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.8216701667148378e-05,
+      "loss": 1.8193,
+      "step": 28900
+    },
+    {
+      "epoch": 1.27,
+      "eval_loss": 1.830406904220581,
+      "eval_runtime": 11.5645,
+      "eval_samples_per_second": 354.187,
+      "eval_steps_per_second": 22.137,
+      "step": 28900
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.819752388116767e-05,
+      "loss": 1.8371,
+      "step": 28910
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.8178351799805637e-05,
+      "loss": 1.8238,
+      "step": 28920
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.8159185432328906e-05,
+      "loss": 1.8504,
+      "step": 28930
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.814002478800136e-05,
+      "loss": 1.7753,
+      "step": 28940
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.812086987608414e-05,
+      "loss": 1.8689,
+      "step": 28950
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.810172070583559e-05,
+      "loss": 1.8385,
+      "step": 28960
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.8082577286511285e-05,
+      "loss": 1.8529,
+      "step": 28970
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.8063439627364016e-05,
+      "loss": 1.8337,
+      "step": 28980
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.804430773764381e-05,
+      "loss": 1.7948,
+      "step": 28990
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.802518162659787e-05,
+      "loss": 1.838,
+      "step": 29000
+    },
+    {
+      "epoch": 1.27,
+      "eval_loss": 1.830385684967041,
+      "eval_runtime": 11.7627,
+      "eval_samples_per_second": 348.218,
+      "eval_steps_per_second": 21.764,
+      "step": 29000
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.8006061303470648e-05,
+      "loss": 1.7526,
+      "step": 29010
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.798694677750376e-05,
+      "loss": 1.8202,
+      "step": 29020
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.796783805793606e-05,
+      "loss": 1.8431,
+      "step": 29030
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.794873515400353e-05,
+      "loss": 1.813,
+      "step": 29040
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.7929638074939433e-05,
+      "loss": 1.8351,
+      "step": 29050
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.7910546829974127e-05,
+      "loss": 1.8675,
+      "step": 29060
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.789146142833522e-05,
+      "loss": 1.8152,
+      "step": 29070
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.787238187924744e-05,
+      "loss": 1.8264,
+      "step": 29080
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.785330819193274e-05,
+      "loss": 1.7769,
+      "step": 29090
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.7834240375610183e-05,
+      "loss": 1.8131,
+      "step": 29100
+    },
+    {
+      "epoch": 1.27,
+      "eval_loss": 1.830368995666504,
+      "eval_runtime": 11.647,
+      "eval_samples_per_second": 351.678,
+      "eval_steps_per_second": 21.98,
+      "step": 29100
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.7815178439496047e-05,
+      "loss": 1.8124,
+      "step": 29110
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.7796122392803727e-05,
+      "loss": 1.8489,
+      "step": 29120
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.77770722447438e-05,
+      "loss": 1.8363,
+      "step": 29130
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.7758028004523978e-05,
+      "loss": 1.8315,
+      "step": 29140
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.773898968134912e-05,
+      "loss": 1.8668,
+      "step": 29150
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.7719957284421228e-05,
+      "loss": 1.8214,
+      "step": 29160
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.7700930822939424e-05,
+      "loss": 1.78,
+      "step": 29170
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.7681910306099985e-05,
+      "loss": 1.8128,
+      "step": 29180
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.766289574309629e-05,
+      "loss": 1.8504,
+      "step": 29190
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.7643887143118875e-05,
+      "loss": 1.7937,
+      "step": 29200
+    },
+    {
+      "epoch": 1.28,
+      "eval_loss": 1.8301746845245361,
+      "eval_runtime": 11.7318,
+      "eval_samples_per_second": 349.135,
+      "eval_steps_per_second": 21.821,
+      "step": 29200
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.7624884515355358e-05,
+      "loss": 1.8176,
+      "step": 29210
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.760588786899049e-05,
+      "loss": 1.8329,
+      "step": 29220
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.758689721320612e-05,
+      "loss": 1.863,
+      "step": 29230
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.7567912557181218e-05,
+      "loss": 1.8621,
+      "step": 29240
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.7548933910091834e-05,
+      "loss": 1.8192,
+      "step": 29250
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.752996128111113e-05,
+      "loss": 1.8057,
+      "step": 29260
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.751099467940934e-05,
+      "loss": 1.8206,
+      "step": 29270
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.7492034114153825e-05,
+      "loss": 1.8019,
+      "step": 29280
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.7473079594508966e-05,
+      "loss": 1.8015,
+      "step": 29290
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.7454131129636273e-05,
+      "loss": 1.8239,
+      "step": 29300
+    },
+    {
+      "epoch": 1.28,
+      "eval_loss": 1.8301327228546143,
+      "eval_runtime": 11.5499,
+      "eval_samples_per_second": 354.636,
+      "eval_steps_per_second": 22.165,
+      "step": 29300
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.7435188728694312e-05,
+      "loss": 1.8463,
+      "step": 29310
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.741625240083873e-05,
+      "loss": 1.8382,
+      "step": 29320
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.7397322155222203e-05,
+      "loss": 1.8513,
+      "step": 29330
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.73783980009945e-05,
+      "loss": 1.7766,
+      "step": 29340
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.735947994730245e-05,
+      "loss": 1.8219,
+      "step": 29350
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.7340568003289917e-05,
+      "loss": 1.851,
+      "step": 29360
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.7321662178097805e-05,
+      "loss": 1.8086,
+      "step": 29370
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.7302762480864093e-05,
+      "loss": 1.7966,
+      "step": 29380
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.7283868920723734e-05,
+      "loss": 1.8585,
+      "step": 29390
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.726498150680881e-05,
+      "loss": 1.8134,
+      "step": 29400
+    },
+    {
+      "epoch": 1.29,
+      "eval_loss": 1.8302505016326904,
+      "eval_runtime": 11.6196,
+      "eval_samples_per_second": 352.509,
+      "eval_steps_per_second": 22.032,
+      "step": 29400
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.7246100248248356e-05,
+      "loss": 1.7909,
+      "step": 29410
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.7227225154168438e-05,
+      "loss": 1.8333,
+      "step": 29420
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.7208356233692174e-05,
+      "loss": 1.7974,
+      "step": 29430
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.718949349593969e-05,
+      "loss": 1.833,
+      "step": 29440
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.717063695002812e-05,
+      "loss": 1.8043,
+      "step": 29450
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.7151786605071588e-05,
+      "loss": 1.8103,
+      "step": 29460
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.713294247018125e-05,
+      "loss": 1.8118,
+      "step": 29470
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.7114104554465216e-05,
+      "loss": 1.8458,
+      "step": 29480
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.709527286702867e-05,
+      "loss": 1.8072,
+      "step": 29490
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.707644741697369e-05,
+      "loss": 1.8057,
+      "step": 29500
+    },
+    {
+      "epoch": 1.29,
+      "eval_loss": 1.8301050662994385,
+      "eval_runtime": 11.921,
+      "eval_samples_per_second": 343.596,
+      "eval_steps_per_second": 21.475,
+      "step": 29500
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.7057628213399415e-05,
+      "loss": 1.8512,
+      "step": 29510
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.703881526540191e-05,
+      "loss": 1.7953,
+      "step": 29520
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.7020008582074257e-05,
+      "loss": 1.8278,
+      "step": 29530
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.7001208172506487e-05,
+      "loss": 1.7863,
+      "step": 29540
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.6982414045785624e-05,
+      "loss": 1.8038,
+      "step": 29550
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.6963626210995608e-05,
+      "loss": 1.861,
+      "step": 29560
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.6944844677217378e-05,
+      "loss": 1.8666,
+      "step": 29570
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.6926069453528822e-05,
+      "loss": 1.8175,
+      "step": 29580
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.690730054900478e-05,
+      "loss": 1.8434,
+      "step": 29590
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.6888537972717008e-05,
+      "loss": 1.8518,
+      "step": 29600
+    },
+    {
+      "epoch": 1.3,
+      "eval_loss": 1.829789400100708,
+      "eval_runtime": 11.5886,
+      "eval_samples_per_second": 353.45,
+      "eval_steps_per_second": 22.091,
+      "step": 29600
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.6869781733734234e-05,
+      "loss": 1.8064,
+      "step": 29610
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.6851031841122114e-05,
+      "loss": 1.8443,
+      "step": 29620
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.6832288303943254e-05,
+      "loss": 1.8501,
+      "step": 29630
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.6813551131257154e-05,
+      "loss": 1.8504,
+      "step": 29640
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.679482033212025e-05,
+      "loss": 1.8269,
+      "step": 29650
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.6776095915585905e-05,
+      "loss": 1.7929,
+      "step": 29660
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.6757377890704398e-05,
+      "loss": 1.8159,
+      "step": 29670
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.6738666266522928e-05,
+      "loss": 1.8405,
+      "step": 29680
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.671996105208556e-05,
+      "loss": 1.7826,
+      "step": 29690
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.6701262256433304e-05,
+      "loss": 1.8001,
+      "step": 29700
+    },
+    {
+      "epoch": 1.3,
+      "eval_loss": 1.8297191858291626,
+      "eval_runtime": 11.5768,
+      "eval_samples_per_second": 353.81,
+      "eval_steps_per_second": 22.113,
+      "step": 29700
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.6682569888604047e-05,
+      "loss": 1.7877,
+      "step": 29710
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.666388395763259e-05,
+      "loss": 1.8405,
+      "step": 29720
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.6645204472550576e-05,
+      "loss": 1.8083,
+      "step": 29730
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.6626531442386598e-05,
+      "loss": 1.7797,
+      "step": 29740
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.6607864876166048e-05,
+      "loss": 1.8313,
+      "step": 29750
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.6589204782911287e-05,
+      "loss": 1.798,
+      "step": 29760
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.6570551171641475e-05,
+      "loss": 1.8381,
+      "step": 29770
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.6551904051372674e-05,
+      "loss": 1.8251,
+      "step": 29780
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.6533263431117786e-05,
+      "loss": 1.7904,
+      "step": 29790
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.6514629319886592e-05,
+      "loss": 1.8533,
+      "step": 29800
+    },
+    {
+      "epoch": 1.3,
+      "eval_loss": 1.8297039270401,
+      "eval_runtime": 11.6268,
+      "eval_samples_per_second": 352.288,
+      "eval_steps_per_second": 22.018,
+      "step": 29800
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.6496001726685715e-05,
+      "loss": 1.8153,
+      "step": 29810
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.647738066051865e-05,
+      "loss": 1.866,
+      "step": 29820
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.6458766130385692e-05,
+      "loss": 1.8188,
+      "step": 29830
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.6440158145284017e-05,
+      "loss": 1.8115,
+      "step": 29840
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.642155671420762e-05,
+      "loss": 1.8042,
+      "step": 29850
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.6402961846147346e-05,
+      "loss": 1.8007,
+      "step": 29860
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.638437355009084e-05,
+      "loss": 1.7936,
+      "step": 29870
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.6365791835022566e-05,
+      "loss": 1.7836,
+      "step": 29880
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.6347216709923872e-05,
+      "loss": 1.7985,
+      "step": 29890
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.6328648183772837e-05,
+      "loss": 1.7926,
+      "step": 29900
+    },
+    {
+      "epoch": 1.31,
+      "eval_loss": 1.82939612865448,
+      "eval_runtime": 11.6247,
+      "eval_samples_per_second": 352.353,
+      "eval_steps_per_second": 22.022,
+      "step": 29900
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.6310086265544412e-05,
+      "loss": 1.8354,
+      "step": 29910
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.6291530964210303e-05,
+      "loss": 1.8529,
+      "step": 29920
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.6272982288739063e-05,
+      "loss": 1.7827,
+      "step": 29930
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.6254440248096022e-05,
+      "loss": 1.7849,
+      "step": 29940
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.623590485124331e-05,
+      "loss": 1.836,
+      "step": 29950
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.621737610713983e-05,
+      "loss": 1.7903,
+      "step": 29960
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.6198854024741286e-05,
+      "loss": 1.7863,
+      "step": 29970
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.6180338613000155e-05,
+      "loss": 1.7768,
+      "step": 29980
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.6161829880865707e-05,
+      "loss": 1.804,
+      "step": 29990
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.6143327837283946e-05,
+      "loss": 1.7666,
+      "step": 30000
+    },
+    {
+      "epoch": 1.31,
+      "eval_loss": 1.8295360803604126,
+      "eval_runtime": 11.8624,
+      "eval_samples_per_second": 345.294,
+      "eval_steps_per_second": 21.581,
+      "step": 30000
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.6124832491197682e-05,
+      "loss": 1.8302,
+      "step": 30010
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.610634385154644e-05,
+      "loss": 1.7949,
+      "step": 30020
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.608786192726658e-05,
+      "loss": 1.8113,
+      "step": 30030
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.606938672729114e-05,
+      "loss": 1.8027,
+      "step": 30040
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.6050918260549955e-05,
+      "loss": 1.7857,
+      "step": 30050
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.6032456535969576e-05,
+      "loss": 1.8274,
+      "step": 30060
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.6014001562473305e-05,
+      "loss": 1.8056,
+      "step": 30070
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.5995553348981197e-05,
+      "loss": 1.7804,
+      "step": 30080
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.5977111904410034e-05,
+      "loss": 1.8346,
+      "step": 30090
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.5958677237673295e-05,
+      "loss": 1.8124,
+      "step": 30100
+    },
+    {
+      "epoch": 1.32,
+      "eval_loss": 1.8036388158798218,
+      "eval_runtime": 12.2283,
+      "eval_samples_per_second": 334.96,
+      "eval_steps_per_second": 20.935,
+      "step": 30100
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.594024935768122e-05,
+      "loss": 1.7914,
+      "step": 30110
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.5921828273340768e-05,
+      "loss": 1.817,
+      "step": 30120
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.590341399355558e-05,
+      "loss": 1.8477,
+      "step": 30130
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.588500652722605e-05,
+      "loss": 1.8133,
+      "step": 30140
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.586660588324923e-05,
+      "loss": 1.8079,
+      "step": 30150
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.5848212070518923e-05,
+      "loss": 1.8397,
+      "step": 30160
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.5829825097925605e-05,
+      "loss": 1.8179,
+      "step": 30170
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.5811444974356466e-05,
+      "loss": 1.8638,
+      "step": 30180
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.579307170869534e-05,
+      "loss": 1.8601,
+      "step": 30190
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.5774705309822796e-05,
+      "loss": 1.8226,
+      "step": 30200
+    },
+    {
+      "epoch": 1.32,
+      "eval_loss": 1.8035895824432373,
+      "eval_runtime": 13.4266,
+      "eval_samples_per_second": 305.066,
+      "eval_steps_per_second": 19.067,
+      "step": 30200
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.575634578661606e-05,
+      "loss": 1.8057,
+      "step": 30210
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.573799314794905e-05,
+      "loss": 1.8171,
+      "step": 30220
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.571964740269233e-05,
+      "loss": 1.8724,
+      "step": 30230
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.570130855971315e-05,
+      "loss": 1.8347,
+      "step": 30240
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.5682976627875423e-05,
+      "loss": 1.8281,
+      "step": 30250
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.566465161603974e-05,
+      "loss": 1.8488,
+      "step": 30260
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.56463335330633e-05,
+      "loss": 1.8638,
+      "step": 30270
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.5628022387799995e-05,
+      "loss": 1.8497,
+      "step": 30280
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.5609718189100322e-05,
+      "loss": 1.796,
+      "step": 30290
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.5591420945811503e-05,
+      "loss": 1.8429,
+      "step": 30300
+    },
+    {
+      "epoch": 1.33,
+      "eval_loss": 1.8034627437591553,
+      "eval_runtime": 11.7516,
+      "eval_samples_per_second": 348.548,
+      "eval_steps_per_second": 21.784,
+      "step": 30300
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.5573130666777293e-05,
+      "loss": 1.8275,
+      "step": 30310
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.5554847360838164e-05,
+      "loss": 1.8428,
+      "step": 30320
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.5536571036831148e-05,
+      "loss": 1.7699,
+      "step": 30330
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.5518301703589967e-05,
+      "loss": 1.7879,
+      "step": 30340
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.550003936994494e-05,
+      "loss": 1.8352,
+      "step": 30350
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.5481784044722975e-05,
+      "loss": 1.8067,
+      "step": 30360
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.5463535736747636e-05,
+      "loss": 1.7945,
+      "step": 30370
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.5445294454839047e-05,
+      "loss": 1.8567,
+      "step": 30380
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.5427060207814008e-05,
+      "loss": 1.8136,
+      "step": 30390
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.540883300448584e-05,
+      "loss": 1.8147,
+      "step": 30400
+    },
+    {
+      "epoch": 1.33,
+      "eval_loss": 1.803378939628601,
+      "eval_runtime": 11.9608,
+      "eval_samples_per_second": 342.451,
+      "eval_steps_per_second": 21.403,
+      "step": 30400
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.5390612853664515e-05,
+      "loss": 1.7916,
+      "step": 30410
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.537239976415656e-05,
+      "loss": 1.7759,
+      "step": 30420
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.5354193744765113e-05,
+      "loss": 1.8284,
+      "step": 30430
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.533599480428988e-05,
+      "loss": 1.7931,
+      "step": 30440
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.5317802951527177e-05,
+      "loss": 1.8196,
+      "step": 30450
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.5299618195269837e-05,
+      "loss": 1.8148,
+      "step": 30460
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.5281440544307304e-05,
+      "loss": 1.835,
+      "step": 30470
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.526327000742559e-05,
+      "loss": 1.7835,
+      "step": 30480
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.5245106593407258e-05,
+      "loss": 1.7868,
+      "step": 30490
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.522695031103141e-05,
+      "loss": 1.8516,
+      "step": 30500
+    },
+    {
+      "epoch": 1.34,
+      "eval_loss": 1.8033866882324219,
+      "eval_runtime": 11.923,
+      "eval_samples_per_second": 343.536,
+      "eval_steps_per_second": 21.471,
+      "step": 30500
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.5208801169073735e-05,
+      "loss": 1.8361,
+      "step": 30510
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.5190659176306442e-05,
+      "loss": 1.8378,
+      "step": 30520
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.5172524341498316e-05,
+      "loss": 1.8046,
+      "step": 30530
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.5154396673414641e-05,
+      "loss": 1.8332,
+      "step": 30540
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.513627618081728e-05,
+      "loss": 1.8318,
+      "step": 30550
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.5118162872464576e-05,
+      "loss": 1.7733,
+      "step": 30560
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.5100056757111473e-05,
+      "loss": 1.8487,
+      "step": 30570
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.5081957843509373e-05,
+      "loss": 1.8064,
+      "step": 30580
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.5063866140406211e-05,
+      "loss": 1.8432,
+      "step": 30590
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.5045781656546451e-05,
+      "loss": 1.823,
+      "step": 30600
+    },
+    {
+      "epoch": 1.34,
+      "eval_loss": 1.8032710552215576,
+      "eval_runtime": 12.014,
+      "eval_samples_per_second": 340.936,
+      "eval_steps_per_second": 21.309,
+      "step": 30600
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.502770440067107e-05,
+      "loss": 1.7994,
+      "step": 30610
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.5009634381517554e-05,
+      "loss": 1.871,
+      "step": 30620
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.4991571607819855e-05,
+      "loss": 1.7873,
+      "step": 30630
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.4973516088308472e-05,
+      "loss": 1.8297,
+      "step": 30640
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.4955467831710347e-05,
+      "loss": 1.803,
+      "step": 30650
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.4937426846748978e-05,
+      "loss": 1.7824,
+      "step": 30660
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.4919393142144276e-05,
+      "loss": 1.8144,
+      "step": 30670
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.4901366726612695e-05,
+      "loss": 1.8307,
+      "step": 30680
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.488334760886711e-05,
+      "loss": 1.8081,
+      "step": 30690
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.4865335797616908e-05,
+      "loss": 1.8277,
+      "step": 30700
+    },
+    {
+      "epoch": 1.34,
+      "eval_loss": 1.8033111095428467,
+      "eval_runtime": 11.7975,
+      "eval_samples_per_second": 347.192,
+      "eval_steps_per_second": 21.699,
+      "step": 30700
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.4847331301567937e-05,
+      "loss": 1.8293,
+      "step": 30710
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.4829334129422513e-05,
+      "loss": 1.8478,
+      "step": 30720
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.4811344289879381e-05,
+      "loss": 1.8669,
+      "step": 30730
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.4793361791633779e-05,
+      "loss": 1.8202,
+      "step": 30740
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.477538664337738e-05,
+      "loss": 1.8127,
+      "step": 30750
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.475741885379832e-05,
+      "loss": 1.8181,
+      "step": 30760
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.4739458431581146e-05,
+      "loss": 1.8136,
+      "step": 30770
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.4721505385406865e-05,
+      "loss": 1.8594,
+      "step": 30780
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.470355972395293e-05,
+      "loss": 1.7994,
+      "step": 30790
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.4685621455893215e-05,
+      "loss": 1.7717,
+      "step": 30800
+    },
+    {
+      "epoch": 1.35,
+      "eval_loss": 1.8033769130706787,
+      "eval_runtime": 12.8899,
+      "eval_samples_per_second": 317.768,
+      "eval_steps_per_second": 19.861,
+      "step": 30800
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.4667690589897995e-05,
+      "loss": 1.8188,
+      "step": 30810
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.4649767134634016e-05,
+      "loss": 1.8189,
+      "step": 30820
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.463185109876439e-05,
+      "loss": 1.8468,
+      "step": 30830
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.4613942490948683e-05,
+      "loss": 1.8815,
+      "step": 30840
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.4596041319842866e-05,
+      "loss": 1.8377,
+      "step": 30850
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.4578147594099282e-05,
+      "loss": 1.8201,
+      "step": 30860
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.4560261322366711e-05,
+      "loss": 1.8229,
+      "step": 30870
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.4542382513290323e-05,
+      "loss": 1.8585,
+      "step": 30880
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.4524511175511686e-05,
+      "loss": 1.7989,
+      "step": 30890
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.4506647317668719e-05,
+      "loss": 1.8391,
+      "step": 30900
+    },
+    {
+      "epoch": 1.35,
+      "eval_loss": 1.8032344579696655,
+      "eval_runtime": 12.1463,
+      "eval_samples_per_second": 337.222,
+      "eval_steps_per_second": 21.076,
+      "step": 30900
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.4488790948395783e-05,
+      "loss": 1.8361,
+      "step": 30910
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.4470942076323553e-05,
+      "loss": 1.8794,
+      "step": 30920
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.4453100710079167e-05,
+      "loss": 1.8314,
+      "step": 30930
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.4435266858286048e-05,
+      "loss": 1.7951,
+      "step": 30940
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.441744052956405e-05,
+      "loss": 1.8046,
+      "step": 30950
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.4399621732529337e-05,
+      "loss": 1.7914,
+      "step": 30960
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.4381810475794482e-05,
+      "loss": 1.7874,
+      "step": 30970
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.4364006767968386e-05,
+      "loss": 1.7981,
+      "step": 30980
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.434621061765632e-05,
+      "loss": 1.8093,
+      "step": 30990
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.432842203345987e-05,
+      "loss": 1.8229,
+      "step": 31000
+    },
+    {
+      "epoch": 1.36,
+      "eval_loss": 1.8032114505767822,
+      "eval_runtime": 12.0468,
+      "eval_samples_per_second": 340.007,
+      "eval_steps_per_second": 21.25,
+      "step": 31000
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.4310641023976996e-05,
+      "loss": 1.8338,
+      "step": 31010
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.4292867597801983e-05,
+      "loss": 1.8277,
+      "step": 31020
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.427510176352547e-05,
+      "loss": 1.8275,
+      "step": 31030
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.425734352973438e-05,
+      "loss": 1.7842,
+      "step": 31040
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.4239592905012024e-05,
+      "loss": 1.8522,
+      "step": 31050
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.4221849897937976e-05,
+      "loss": 1.8172,
+      "step": 31060
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.4204114517088168e-05,
+      "loss": 1.7892,
+      "step": 31070
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.4186386771034842e-05,
+      "loss": 1.8359,
+      "step": 31080
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.4168666668346526e-05,
+      "loss": 1.8521,
+      "step": 31090
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.4150954217588076e-05,
+      "loss": 1.7878,
+      "step": 31100
+    },
+    {
+      "epoch": 1.36,
+      "eval_loss": 1.8032019138336182,
+      "eval_runtime": 18.0994,
+      "eval_samples_per_second": 226.306,
+      "eval_steps_per_second": 14.144,
+      "step": 31100
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.4133249427320644e-05,
+      "loss": 1.8111,
+      "step": 31110
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.4115552306101688e-05,
+      "loss": 1.8599,
+      "step": 31120
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.4097862862484926e-05,
+      "loss": 1.7708,
+      "step": 31130
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.4080181105020406e-05,
+      "loss": 1.79,
+      "step": 31140
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.4062507042254434e-05,
+      "loss": 1.8523,
+      "step": 31150
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.4044840682729622e-05,
+      "loss": 1.7994,
+      "step": 31160
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.4027182034984823e-05,
+      "loss": 1.8007,
+      "step": 31170
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.4009531107555202e-05,
+      "loss": 1.7724,
+      "step": 31180
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.3991887908972142e-05,
+      "loss": 1.7872,
+      "step": 31190
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.397425244776336e-05,
+      "loss": 1.8078,
+      "step": 31200
+    },
+    {
+      "epoch": 1.37,
+      "eval_loss": 1.80314040184021,
+      "eval_runtime": 12.2739,
+      "eval_samples_per_second": 333.717,
+      "eval_steps_per_second": 20.857,
+      "step": 31200
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.3956624732452768e-05,
+      "loss": 1.815,
+      "step": 31210
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.3939004771560581e-05,
+      "loss": 1.8561,
+      "step": 31220
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.392139257360322e-05,
+      "loss": 1.7922,
+      "step": 31230
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.3903788147093393e-05,
+      "loss": 1.7885,
+      "step": 31240
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.3886191500540042e-05,
+      "loss": 1.7975,
+      "step": 31250
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.386860264244835e-05,
+      "loss": 1.8109,
+      "step": 31260
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.3851021581319709e-05,
+      "loss": 1.7984,
+      "step": 31270
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.3833448325651776e-05,
+      "loss": 1.8225,
+      "step": 31280
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.3815882883938435e-05,
+      "loss": 1.7602,
+      "step": 31290
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.3798325264669757e-05,
+      "loss": 1.786,
+      "step": 31300
+    },
+    {
+      "epoch": 1.37,
+      "eval_loss": 1.802952766418457,
+      "eval_runtime": 15.4326,
+      "eval_samples_per_second": 265.413,
+      "eval_steps_per_second": 16.588,
+      "step": 31300
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.3780775476332083e-05,
+      "loss": 1.8364,
+      "step": 31310
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.3763233527407915e-05,
+      "loss": 1.8026,
+      "step": 31320
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.3745699426376008e-05,
+      "loss": 1.803,
+      "step": 31330
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.3728173181711306e-05,
+      "loss": 1.808,
+      "step": 31340
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.3710654801884973e-05,
+      "loss": 1.7936,
+      "step": 31350
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.3693144295364332e-05,
+      "loss": 1.8277,
+      "step": 31360
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.3675641670612937e-05,
+      "loss": 1.802,
+      "step": 31370
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.3658146936090526e-05,
+      "loss": 1.8276,
+      "step": 31380
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.3640660100253026e-05,
+      "loss": 1.7541,
+      "step": 31390
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.3623181171552512e-05,
+      "loss": 1.8347,
+      "step": 31400
+    },
+    {
+      "epoch": 1.37,
+      "eval_loss": 1.8029532432556152,
+      "eval_runtime": 22.8254,
+      "eval_samples_per_second": 179.449,
+      "eval_steps_per_second": 11.216,
+      "step": 31400
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.3605710158437284e-05,
+      "loss": 1.8161,
+      "step": 31410
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.3588247069351793e-05,
+      "loss": 1.8135,
+      "step": 31420
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.357079191273667e-05,
+      "loss": 1.8231,
+      "step": 31430
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.3553344697028687e-05,
+      "loss": 1.8008,
+      "step": 31440
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.353590543066082e-05,
+      "loss": 1.8755,
+      "step": 31450
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.3518474122062139e-05,
+      "loss": 1.7891,
+      "step": 31460
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.3501050779657955e-05,
+      "loss": 1.7602,
+      "step": 31470
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.3483635411869648e-05,
+      "loss": 1.7987,
+      "step": 31480
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.34662280271148e-05,
+      "loss": 1.8161,
+      "step": 31490
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.3448828633807086e-05,
+      "loss": 1.7987,
+      "step": 31500
+    },
+    {
+      "epoch": 1.38,
+      "eval_loss": 1.8029508590698242,
+      "eval_runtime": 13.1282,
+      "eval_samples_per_second": 311.999,
+      "eval_steps_per_second": 19.5,
+      "step": 31500
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.343143724035635e-05,
+      "loss": 1.8132,
+      "step": 31510
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.3414053855168574e-05,
+      "loss": 1.7628,
+      "step": 31520
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.3396678486645861e-05,
+      "loss": 1.8087,
+      "step": 31530
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.3379311143186428e-05,
+      "loss": 1.8384,
+      "step": 31540
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.3361951833184595e-05,
+      "loss": 1.8143,
+      "step": 31550
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.3344600565030876e-05,
+      "loss": 1.784,
+      "step": 31560
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.3327257347111809e-05,
+      "loss": 1.8613,
+      "step": 31570
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.3309922187810103e-05,
+      "loss": 1.817,
+      "step": 31580
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.3292595095504527e-05,
+      "loss": 1.8067,
+      "step": 31590
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.3275276078569982e-05,
+      "loss": 1.8025,
+      "step": 31600
+    },
+    {
+      "epoch": 1.38,
+      "eval_loss": 1.8028916120529175,
+      "eval_runtime": 12.1557,
+      "eval_samples_per_second": 336.961,
+      "eval_steps_per_second": 21.06,
+      "step": 31600
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.3257965145377457e-05,
+      "loss": 1.8405,
+      "step": 31610
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.3240662304294046e-05,
+      "loss": 1.8196,
+      "step": 31620
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.3223367563682894e-05,
+      "loss": 1.8064,
+      "step": 31630
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.3206080931903264e-05,
+      "loss": 1.8144,
+      "step": 31640
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.3188802417310497e-05,
+      "loss": 1.7973,
+      "step": 31650
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.3171532028256013e-05,
+      "loss": 1.8425,
+      "step": 31660
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.3154269773087273e-05,
+      "loss": 1.7956,
+      "step": 31670
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.313701566014784e-05,
+      "loss": 1.8636,
+      "step": 31680
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.3119769697777336e-05,
+      "loss": 1.8104,
+      "step": 31690
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.3102531894311445e-05,
+      "loss": 1.8135,
+      "step": 31700
+    },
+    {
+      "epoch": 1.39,
+      "eval_loss": 1.8029296398162842,
+      "eval_runtime": 13.8361,
+      "eval_samples_per_second": 296.037,
+      "eval_steps_per_second": 18.502,
+      "step": 31700
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.3085302258081881e-05,
+      "loss": 1.8348,
+      "step": 31710
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.3068080797416454e-05,
+      "loss": 1.828,
+      "step": 31720
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.3050867520638964e-05,
+      "loss": 1.8008,
+      "step": 31730
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.3033662436069339e-05,
+      "loss": 1.8172,
+      "step": 31740
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.3016465552023458e-05,
+      "loss": 1.839,
+      "step": 31750
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.2999276876813308e-05,
+      "loss": 1.8226,
+      "step": 31760
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.298209641874685e-05,
+      "loss": 1.8308,
+      "step": 31770
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.2964924186128115e-05,
+      "loss": 1.8015,
+      "step": 31780
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.2947760187257149e-05,
+      "loss": 1.792,
+      "step": 31790
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.2930604430430003e-05,
+      "loss": 1.8778,
+      "step": 31800
+    },
+    {
+      "epoch": 1.39,
+      "eval_loss": 1.8028862476348877,
+      "eval_runtime": 12.2991,
+      "eval_samples_per_second": 333.031,
+      "eval_steps_per_second": 20.814,
+      "step": 31800
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.2913456923938766e-05,
+      "loss": 1.8518,
+      "step": 31810
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.2896317676071497e-05,
+      "loss": 1.8393,
+      "step": 31820
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.2879186695112344e-05,
+      "loss": 1.806,
+      "step": 31830
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.2862063989341374e-05,
+      "loss": 1.8155,
+      "step": 31840
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.2844949567034711e-05,
+      "loss": 1.8201,
+      "step": 31850
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.282784343646443e-05,
+      "loss": 1.8248,
+      "step": 31860
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.2810745605898638e-05,
+      "loss": 1.7913,
+      "step": 31870
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.279365608360141e-05,
+      "loss": 1.789,
+      "step": 31880
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.2776574877832827e-05,
+      "loss": 1.7382,
+      "step": 31890
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.2759501996848903e-05,
+      "loss": 1.815,
+      "step": 31900
+    },
+    {
+      "epoch": 1.4,
+      "eval_loss": 1.8028854131698608,
+      "eval_runtime": 11.9394,
+      "eval_samples_per_second": 343.066,
+      "eval_steps_per_second": 21.442,
+      "step": 31900
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.2742437448901678e-05,
+      "loss": 1.7857,
+      "step": 31910
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.2725381242239139e-05,
+      "loss": 1.791,
+      "step": 31920
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.2708333385105263e-05,
+      "loss": 1.8129,
+      "step": 31930
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.2691293885739948e-05,
+      "loss": 1.8506,
+      "step": 31940
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.2674262752379095e-05,
+      "loss": 1.8381,
+      "step": 31950
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.2657239993254544e-05,
+      "loss": 1.7665,
+      "step": 31960
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.2640225616594102e-05,
+      "loss": 1.8019,
+      "step": 31970
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.262321963062149e-05,
+      "loss": 1.823,
+      "step": 31980
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.2606222043556413e-05,
+      "loss": 1.8168,
+      "step": 31990
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.2589232863614483e-05,
+      "loss": 1.8158,
+      "step": 32000
+    },
+    {
+      "epoch": 1.4,
+      "eval_loss": 1.802783727645874,
+      "eval_runtime": 11.8236,
+      "eval_samples_per_second": 346.426,
+      "eval_steps_per_second": 21.652,
+      "step": 32000
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.2572252099007274e-05,
+      "loss": 1.8244,
+      "step": 32010
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.2555279757942292e-05,
+      "loss": 1.8271,
+      "step": 32020
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.2538315848622944e-05,
+      "loss": 1.8189,
+      "step": 32030
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.2521360379248586e-05,
+      "loss": 1.805,
+      "step": 32040
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.2504413358014493e-05,
+      "loss": 1.798,
+      "step": 32050
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.248747479311186e-05,
+      "loss": 1.8368,
+      "step": 32060
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.2470544692727772e-05,
+      "loss": 1.8545,
+      "step": 32070
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.2453623065045249e-05,
+      "loss": 1.8446,
+      "step": 32080
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.2436709918243177e-05,
+      "loss": 1.8307,
+      "step": 32090
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.2419805260496419e-05,
+      "loss": 1.8157,
+      "step": 32100
+    },
+    {
+      "epoch": 1.41,
+      "eval_loss": 1.8028050661087036,
+      "eval_runtime": 12.0893,
+      "eval_samples_per_second": 338.811,
+      "eval_steps_per_second": 21.176,
+      "step": 32100
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.2402909099975648e-05,
+      "loss": 1.8266,
+      "step": 32110
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.2386021444847493e-05,
+      "loss": 1.821,
+      "step": 32120
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.236914230327442e-05,
+      "loss": 1.8437,
+      "step": 32130
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.2352271683414824e-05,
+      "loss": 1.7839,
+      "step": 32140
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.2335409593422967e-05,
+      "loss": 1.7678,
+      "step": 32150
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.2318556041448987e-05,
+      "loss": 1.8096,
+      "step": 32160
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.2301711035638879e-05,
+      "loss": 1.8305,
+      "step": 32170
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.228487458413453e-05,
+      "loss": 1.8427,
+      "step": 32180
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.2268046695073689e-05,
+      "loss": 1.8266,
+      "step": 32190
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.225122737658997e-05,
+      "loss": 1.8164,
+      "step": 32200
+    },
+    {
+      "epoch": 1.41,
+      "eval_loss": 1.802794337272644,
+      "eval_runtime": 12.9249,
+      "eval_samples_per_second": 316.908,
+      "eval_steps_per_second": 19.807,
+      "step": 32200
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.2234416636812812e-05,
+      "loss": 1.7873,
+      "step": 32210
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.2217614483867556e-05,
+      "loss": 1.863,
+      "step": 32220
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.220082092587534e-05,
+      "loss": 1.8111,
+      "step": 32230
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.2184035970953213e-05,
+      "loss": 1.8467,
+      "step": 32240
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.216725962721401e-05,
+      "loss": 1.8348,
+      "step": 32250
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.2150491902766414e-05,
+      "loss": 1.8502,
+      "step": 32260
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.2133732805714963e-05,
+      "loss": 1.7893,
+      "step": 32270
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.2116982344160005e-05,
+      "loss": 1.8066,
+      "step": 32280
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.2100240526197742e-05,
+      "loss": 1.835,
+      "step": 32290
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.2083507359920152e-05,
+      "loss": 1.8376,
+      "step": 32300
+    },
+    {
+      "epoch": 1.41,
+      "eval_loss": 1.8027287721633911,
+      "eval_runtime": 11.923,
+      "eval_samples_per_second": 343.538,
+      "eval_steps_per_second": 21.471,
+      "step": 32300
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.2066782853415066e-05,
+      "loss": 1.775,
+      "step": 32310
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.2050067014766129e-05,
+      "loss": 1.7925,
+      "step": 32320
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.2033359852052793e-05,
+      "loss": 1.8244,
+      "step": 32330
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.2016661373350291e-05,
+      "loss": 1.7604,
+      "step": 32340
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1999971586729705e-05,
+      "loss": 1.8296,
+      "step": 32350
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1983290500257852e-05,
+      "loss": 1.7692,
+      "step": 32360
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1966618121997428e-05,
+      "loss": 1.8503,
+      "step": 32370
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1949954460006843e-05,
+      "loss": 1.8422,
+      "step": 32380
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1933299522340345e-05,
+      "loss": 1.8154,
+      "step": 32390
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1916653317047927e-05,
+      "loss": 1.8168,
+      "step": 32400
+    },
+    {
+      "epoch": 1.42,
+      "eval_loss": 1.8027868270874023,
+      "eval_runtime": 12.029,
+      "eval_samples_per_second": 340.51,
+      "eval_steps_per_second": 21.282,
+      "step": 32400
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1900015852175388e-05,
+      "loss": 1.7937,
+      "step": 32410
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1883387135764296e-05,
+      "loss": 1.822,
+      "step": 32420
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1866767175851998e-05,
+      "loss": 1.8388,
+      "step": 32430
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1850155980471576e-05,
+      "loss": 1.7809,
+      "step": 32440
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1833553557651913e-05,
+      "loss": 1.797,
+      "step": 32450
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1816959915417634e-05,
+      "loss": 1.841,
+      "step": 32460
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1800375061789142e-05,
+      "loss": 1.7972,
+      "step": 32470
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1783799004782554e-05,
+      "loss": 1.8139,
+      "step": 32480
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1767231752409753e-05,
+      "loss": 1.8422,
+      "step": 32490
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1750673312678374e-05,
+      "loss": 1.8647,
+      "step": 32500
+    },
+    {
+      "epoch": 1.42,
+      "eval_loss": 1.802736759185791,
+      "eval_runtime": 11.754,
+      "eval_samples_per_second": 348.478,
+      "eval_steps_per_second": 21.78,
+      "step": 32500
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.173412369359179e-05,
+      "loss": 1.8132,
+      "step": 32510
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1717582903149114e-05,
+      "loss": 1.7861,
+      "step": 32520
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1701050949345166e-05,
+      "loss": 1.8471,
+      "step": 32530
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1684527840170532e-05,
+      "loss": 1.816,
+      "step": 32540
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1668013583611494e-05,
+      "loss": 1.8111,
+      "step": 32550
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1651508187650084e-05,
+      "loss": 1.7778,
+      "step": 32560
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1635011660264008e-05,
+      "loss": 1.8028,
+      "step": 32570
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.161852400942673e-05,
+      "loss": 1.8269,
+      "step": 32580
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1602045243107399e-05,
+      "loss": 1.8296,
+      "step": 32590
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1585575369270884e-05,
+      "loss": 1.8497,
+      "step": 32600
+    },
+    {
+      "epoch": 1.43,
+      "eval_loss": 1.797649621963501,
+      "eval_runtime": 12.0427,
+      "eval_samples_per_second": 340.123,
+      "eval_steps_per_second": 21.258,
+      "step": 32600
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1569114395877735e-05,
+      "loss": 1.7884,
+      "step": 32610
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1552662330884225e-05,
+      "loss": 1.8655,
+      "step": 32620
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1536219182242281e-05,
+      "loss": 1.8085,
+      "step": 32630
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1519784957899591e-05,
+      "loss": 1.8425,
+      "step": 32640
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1503359665799452e-05,
+      "loss": 1.7779,
+      "step": 32650
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1486943313880902e-05,
+      "loss": 1.8367,
+      "step": 32660
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1470535910078613e-05,
+      "loss": 1.8409,
+      "step": 32670
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1454137462322965e-05,
+      "loss": 1.7905,
+      "step": 32680
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1437747978539995e-05,
+      "loss": 1.8579,
+      "step": 32690
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1421367466651418e-05,
+      "loss": 1.8186,
+      "step": 32700
+    },
+    {
+      "epoch": 1.43,
+      "eval_loss": 1.7975997924804688,
+      "eval_runtime": 12.4174,
+      "eval_samples_per_second": 329.86,
+      "eval_steps_per_second": 20.616,
+      "step": 32700
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.14049959345746e-05,
+      "loss": 1.818,
+      "step": 32710
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.138863339022255e-05,
+      "loss": 1.8632,
+      "step": 32720
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1372279841503986e-05,
+      "loss": 1.7715,
+      "step": 32730
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1355935296323226e-05,
+      "loss": 1.8207,
+      "step": 32740
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.133959976258027e-05,
+      "loss": 1.825,
+      "step": 32750
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1323273248170733e-05,
+      "loss": 1.8418,
+      "step": 32760
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1306955760985888e-05,
+      "loss": 1.8071,
+      "step": 32770
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1290647308912653e-05,
+      "loss": 1.8258,
+      "step": 32780
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1274347899833573e-05,
+      "loss": 1.8413,
+      "step": 32790
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.12580575416268e-05,
+      "loss": 1.8404,
+      "step": 32800
+    },
+    {
+      "epoch": 1.44,
+      "eval_loss": 1.797699213027954,
+      "eval_runtime": 13.4742,
+      "eval_samples_per_second": 303.988,
+      "eval_steps_per_second": 18.999,
+      "step": 32800
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1241776242166136e-05,
+      "loss": 1.8321,
+      "step": 32810
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1225504009321002e-05,
+      "loss": 1.8252,
+      "step": 32820
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.120924085095644e-05,
+      "loss": 1.7925,
+      "step": 32830
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1192986774933079e-05,
+      "loss": 1.7826,
+      "step": 32840
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1176741789107188e-05,
+      "loss": 1.8035,
+      "step": 32850
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1160505901330634e-05,
+      "loss": 1.8092,
+      "step": 32860
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.114427911945089e-05,
+      "loss": 1.8353,
+      "step": 32870
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1128061451311007e-05,
+      "loss": 1.7647,
+      "step": 32880
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1111852904749665e-05,
+      "loss": 1.8204,
+      "step": 32890
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1095653487601097e-05,
+      "loss": 1.7997,
+      "step": 32900
+    },
+    {
+      "epoch": 1.44,
+      "eval_loss": 1.7977631092071533,
+      "eval_runtime": 11.7853,
+      "eval_samples_per_second": 347.552,
+      "eval_steps_per_second": 21.722,
+      "step": 32900
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1079463207695147e-05,
+      "loss": 1.7835,
+      "step": 32910
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1063282072857246e-05,
+      "loss": 1.8071,
+      "step": 32920
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1047110090908403e-05,
+      "loss": 1.8531,
+      "step": 32930
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1030947269665177e-05,
+      "loss": 1.8684,
+      "step": 32940
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1014793616939731e-05,
+      "loss": 1.8099,
+      "step": 32950
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.0998649140539795e-05,
+      "loss": 1.8266,
+      "step": 32960
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.0982513848268635e-05,
+      "loss": 1.8116,
+      "step": 32970
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.0966387747925117e-05,
+      "loss": 1.7971,
+      "step": 32980
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.0950270847303609e-05,
+      "loss": 1.7828,
+      "step": 32990
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.093416315419411e-05,
+      "loss": 1.8123,
+      "step": 33000
+    },
+    {
+      "epoch": 1.44,
+      "eval_loss": 1.7976471185684204,
+      "eval_runtime": 11.5471,
+      "eval_samples_per_second": 354.72,
+      "eval_steps_per_second": 22.17,
+      "step": 33000
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.0918064676382096e-05,
+      "loss": 1.8024,
+      "step": 33010
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.090197542164864e-05,
+      "loss": 1.7934,
+      "step": 33020
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.0885895397770312e-05,
+      "loss": 1.8409,
+      "step": 33030
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.086982461251926e-05,
+      "loss": 1.8419,
+      "step": 33040
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.0853763073663145e-05,
+      "loss": 1.871,
+      "step": 33050
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.0837710788965178e-05,
+      "loss": 1.8298,
+      "step": 33060
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.0821667766184062e-05,
+      "loss": 1.8285,
+      "step": 33070
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.0805634013074057e-05,
+      "loss": 1.7954,
+      "step": 33080
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.0789609537384931e-05,
+      "loss": 1.8158,
+      "step": 33090
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.0773594346861972e-05,
+      "loss": 1.8429,
+      "step": 33100
+    },
+    {
+      "epoch": 1.45,
+      "eval_loss": 1.797677755355835,
+      "eval_runtime": 11.8947,
+      "eval_samples_per_second": 344.354,
+      "eval_steps_per_second": 21.522,
+      "step": 33100
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.0757588449245962e-05,
+      "loss": 1.804,
+      "step": 33110
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.074159185227321e-05,
+      "loss": 1.8201,
+      "step": 33120
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.0725604563675529e-05,
+      "loss": 1.8097,
+      "step": 33130
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.0709626591180235e-05,
+      "loss": 1.7624,
+      "step": 33140
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.0693657942510116e-05,
+      "loss": 1.8371,
+      "step": 33150
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.067769862538349e-05,
+      "loss": 1.8046,
+      "step": 33160
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.066174864751413e-05,
+      "loss": 1.8491,
+      "step": 33170
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.064580801661132e-05,
+      "loss": 1.8462,
+      "step": 33180
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.0629876740379831e-05,
+      "loss": 1.7958,
+      "step": 33190
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.0613954826519874e-05,
+      "loss": 1.8365,
+      "step": 33200
+    },
+    {
+      "epoch": 1.45,
+      "eval_loss": 1.7976925373077393,
+      "eval_runtime": 11.8876,
+      "eval_samples_per_second": 344.56,
+      "eval_steps_per_second": 21.535,
+      "step": 33200
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.059804228272717e-05,
+      "loss": 1.8233,
+      "step": 33210
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.0582139116692907e-05,
+      "loss": 1.8124,
+      "step": 33220
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.0566245336103738e-05,
+      "loss": 1.8068,
+      "step": 33230
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.055036094864176e-05,
+      "loss": 1.8066,
+      "step": 33240
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0534485961984566e-05,
+      "loss": 1.7826,
+      "step": 33250
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0518620383805156e-05,
+      "loss": 1.8468,
+      "step": 33260
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0502764221772047e-05,
+      "loss": 1.8527,
+      "step": 33270
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0486917483549145e-05,
+      "loss": 1.8265,
+      "step": 33280
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0471080176795845e-05,
+      "loss": 1.7898,
+      "step": 33290
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0455252309166941e-05,
+      "loss": 1.8109,
+      "step": 33300
+    },
+    {
+      "epoch": 1.46,
+      "eval_loss": 1.7976725101470947,
+      "eval_runtime": 11.8849,
+      "eval_samples_per_second": 344.639,
+      "eval_steps_per_second": 21.54,
+      "step": 33300
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0439433888312702e-05,
+      "loss": 1.8551,
+      "step": 33310
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0423624921878813e-05,
+      "loss": 1.8199,
+      "step": 33320
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0407825417506404e-05,
+      "loss": 1.8212,
+      "step": 33330
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0392035382832e-05,
+      "loss": 1.8219,
+      "step": 33340
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0376254825487582e-05,
+      "loss": 1.75,
+      "step": 33350
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0360483753100536e-05,
+      "loss": 1.8406,
+      "step": 33360
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0344722173293676e-05,
+      "loss": 1.8288,
+      "step": 33370
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0328970093685193e-05,
+      "loss": 1.8457,
+      "step": 33380
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0313227521888739e-05,
+      "loss": 1.858,
+      "step": 33390
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0297494465513307e-05,
+      "loss": 1.8263,
+      "step": 33400
+    },
+    {
+      "epoch": 1.46,
+      "eval_loss": 1.7975435256958008,
+      "eval_runtime": 11.9333,
+      "eval_samples_per_second": 343.24,
+      "eval_steps_per_second": 21.453,
+      "step": 33400
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0281770932163362e-05,
+      "loss": 1.759,
+      "step": 33410
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0266056929438715e-05,
+      "loss": 1.8427,
+      "step": 33420
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0250352464934572e-05,
+      "loss": 1.8069,
+      "step": 33430
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0234657546241556e-05,
+      "loss": 1.7832,
+      "step": 33440
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0218972180945653e-05,
+      "loss": 1.8104,
+      "step": 33450
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0203296376628262e-05,
+      "loss": 1.8281,
+      "step": 33460
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.0187630140866115e-05,
+      "loss": 1.7494,
+      "step": 33470
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.0171973481231352e-05,
+      "loss": 1.8211,
+      "step": 33480
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.0156326405291476e-05,
+      "loss": 1.8315,
+      "step": 33490
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.014068892060937e-05,
+      "loss": 1.7995,
+      "step": 33500
+    },
+    {
+      "epoch": 1.47,
+      "eval_loss": 1.7975175380706787,
+      "eval_runtime": 11.6479,
+      "eval_samples_per_second": 351.65,
+      "eval_steps_per_second": 21.978,
+      "step": 33500
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.0125061034743244e-05,
+      "loss": 1.7929,
+      "step": 33510
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.0109442755246719e-05,
+      "loss": 1.8212,
+      "step": 33520
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.0093834089668715e-05,
+      "loss": 1.8553,
+      "step": 33530
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.0078235045553576e-05,
+      "loss": 1.7782,
+      "step": 33540
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.0062645630440931e-05,
+      "loss": 1.8372,
+      "step": 33550
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.0047065851865797e-05,
+      "loss": 1.8232,
+      "step": 33560
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.0031495717358497e-05,
+      "loss": 1.8573,
+      "step": 33570
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.0015935234444728e-05,
+      "loss": 1.8615,
+      "step": 33580
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.00003844106455e-05,
+      "loss": 1.7935,
+      "step": 33590
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 9.984843253477175e-06,
+      "loss": 1.7813,
+      "step": 33600
+    },
+    {
+      "epoch": 1.47,
+      "eval_loss": 1.7975555658340454,
+      "eval_runtime": 11.8662,
+      "eval_samples_per_second": 345.182,
+      "eval_steps_per_second": 21.574,
+      "step": 33600
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 9.96931177045141e-06,
+      "loss": 1.8015,
+      "step": 33610
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 9.95378996907521e-06,
+      "loss": 1.8304,
+      "step": 33620
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 9.938277856850903e-06,
+      "loss": 1.8519,
+      "step": 33630
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 9.922775441276128e-06,
+      "loss": 1.7916,
+      "step": 33640
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 9.907282729843818e-06,
+      "loss": 1.8084,
+      "step": 33650
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 9.891799730042259e-06,
+      "loss": 1.8415,
+      "step": 33660
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 9.87632644935499e-06,
+      "loss": 1.8212,
+      "step": 33670
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 9.860862895260891e-06,
+      "loss": 1.8296,
+      "step": 33680
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 9.845409075234143e-06,
+      "loss": 1.8163,
+      "step": 33690
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 9.829964996744183e-06,
+      "loss": 1.7989,
+      "step": 33700
+    },
+    {
+      "epoch": 1.48,
+      "eval_loss": 1.7975475788116455,
+      "eval_runtime": 11.8379,
+      "eval_samples_per_second": 346.008,
+      "eval_steps_per_second": 21.626,
+      "step": 33700
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 9.814530667255777e-06,
+      "loss": 1.8075,
+      "step": 33710
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 9.799106094228968e-06,
+      "loss": 1.8176,
+      "step": 33720
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 9.783691285119089e-06,
+      "loss": 1.8055,
+      "step": 33730
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 9.768286247376732e-06,
+      "loss": 1.7935,
+      "step": 33740
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 9.752890988447783e-06,
+      "loss": 1.7922,
+      "step": 33750
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 9.737505515773408e-06,
+      "loss": 1.8302,
+      "step": 33760
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 9.722129836790034e-06,
+      "loss": 1.8262,
+      "step": 33770
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 9.706763958929344e-06,
+      "loss": 1.8267,
+      "step": 33780
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 9.691407889618314e-06,
+      "loss": 1.8011,
+      "step": 33790
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 9.67606163627912e-06,
+      "loss": 1.7894,
+      "step": 33800
+    },
+    {
+      "epoch": 1.48,
+      "eval_loss": 1.7975398302078247,
+      "eval_runtime": 11.7714,
+      "eval_samples_per_second": 347.963,
+      "eval_steps_per_second": 21.748,
+      "step": 33800
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 9.660725206329281e-06,
+      "loss": 1.8095,
+      "step": 33810
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 9.645398607181486e-06,
+      "loss": 1.8024,
+      "step": 33820
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 9.630081846243732e-06,
+      "loss": 1.857,
+      "step": 33830
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 9.614774930919208e-06,
+      "loss": 1.8129,
+      "step": 33840
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 9.599477868606384e-06,
+      "loss": 1.8335,
+      "step": 33850
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 9.584190666698953e-06,
+      "loss": 1.7774,
+      "step": 33860
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 9.56891333258586e-06,
+      "loss": 1.7853,
+      "step": 33870
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 9.553645873651238e-06,
+      "loss": 1.7977,
+      "step": 33880
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 9.538388297274485e-06,
+      "loss": 1.856,
+      "step": 33890
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 9.523140610830224e-06,
+      "loss": 1.733,
+      "step": 33900
+    },
+    {
+      "epoch": 1.48,
+      "eval_loss": 1.7975311279296875,
+      "eval_runtime": 11.7099,
+      "eval_samples_per_second": 349.789,
+      "eval_steps_per_second": 21.862,
+      "step": 33900
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 9.507902821688261e-06,
+      "loss": 1.8348,
+      "step": 33910
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 9.492674937213659e-06,
+      "loss": 1.819,
+      "step": 33920
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 9.477456964766666e-06,
+      "loss": 1.8414,
+      "step": 33930
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 9.46224891170275e-06,
+      "loss": 1.8223,
+      "step": 33940
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 9.447050785372585e-06,
+      "loss": 1.8267,
+      "step": 33950
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 9.431862593122065e-06,
+      "loss": 1.8111,
+      "step": 33960
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 9.416684342292233e-06,
+      "loss": 1.7722,
+      "step": 33970
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 9.40151604021937e-06,
+      "loss": 1.8534,
+      "step": 33980
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 9.38635769423494e-06,
+      "loss": 1.8012,
+      "step": 33990
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 9.371209311665597e-06,
+      "loss": 1.788,
+      "step": 34000
+    },
+    {
+      "epoch": 1.49,
+      "eval_loss": 1.7975867986679077,
+      "eval_runtime": 11.6503,
+      "eval_samples_per_second": 351.579,
+      "eval_steps_per_second": 21.974,
+      "step": 34000
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 9.356070899833157e-06,
+      "loss": 1.7657,
+      "step": 34010
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 9.340942466054636e-06,
+      "loss": 1.8549,
+      "step": 34020
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 9.32582401764223e-06,
+      "loss": 1.8214,
+      "step": 34030
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 9.310715561903305e-06,
+      "loss": 1.8474,
+      "step": 34040
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 9.295617106140383e-06,
+      "loss": 1.7682,
+      "step": 34050
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 9.28052865765118e-06,
+      "loss": 1.895,
+      "step": 34060
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 9.265450223728538e-06,
+      "loss": 1.8382,
+      "step": 34070
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 9.250381811660492e-06,
+      "loss": 1.8462,
+      "step": 34080
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 9.235323428730223e-06,
+      "loss": 1.847,
+      "step": 34090
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 9.220275082216066e-06,
+      "loss": 1.8147,
+      "step": 34100
+    },
+    {
+      "epoch": 1.49,
+      "eval_loss": 1.7975001335144043,
+      "eval_runtime": 11.9247,
+      "eval_samples_per_second": 343.489,
+      "eval_steps_per_second": 21.468,
+      "step": 34100
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 9.20523677939149e-06,
+      "loss": 1.8361,
+      "step": 34110
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 9.190208527525124e-06,
+      "loss": 1.7882,
+      "step": 34120
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 9.17519033388074e-06,
+      "loss": 1.7995,
+      "step": 34130
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 9.160182205717249e-06,
+      "loss": 1.8362,
+      "step": 34140
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 9.145184150288683e-06,
+      "loss": 1.8442,
+      "step": 34150
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 9.130196174844202e-06,
+      "loss": 1.7843,
+      "step": 34160
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 9.11521828662814e-06,
+      "loss": 1.8043,
+      "step": 34170
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 9.100250492879893e-06,
+      "loss": 1.7229,
+      "step": 34180
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 9.085292800834028e-06,
+      "loss": 1.8089,
+      "step": 34190
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 9.070345217720187e-06,
+      "loss": 1.8034,
+      "step": 34200
+    },
+    {
+      "epoch": 1.5,
+      "eval_loss": 1.7975050210952759,
+      "eval_runtime": 11.8339,
+      "eval_samples_per_second": 346.124,
+      "eval_steps_per_second": 21.633,
+      "step": 34200
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 9.055407750763159e-06,
+      "loss": 1.8414,
+      "step": 34210
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 9.040480407182829e-06,
+      "loss": 1.833,
+      "step": 34220
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 9.025563194194207e-06,
+      "loss": 1.8378,
+      "step": 34230
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 9.010656119007366e-06,
+      "loss": 1.8454,
+      "step": 34240
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 8.995759188827512e-06,
+      "loss": 1.8588,
+      "step": 34250
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 8.980872410854948e-06,
+      "loss": 1.8293,
+      "step": 34260
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 8.965995792285069e-06,
+      "loss": 1.7856,
+      "step": 34270
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 8.951129340308327e-06,
+      "loss": 1.8215,
+      "step": 34280
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 8.936273062110308e-06,
+      "loss": 1.7872,
+      "step": 34290
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 8.921426964871632e-06,
+      "loss": 1.8376,
+      "step": 34300
+    },
+    {
+      "epoch": 1.5,
+      "eval_loss": 1.7974214553833008,
+      "eval_runtime": 11.8688,
+      "eval_samples_per_second": 345.108,
+      "eval_steps_per_second": 21.569,
+      "step": 34300
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 8.906591055768057e-06,
+      "loss": 1.7878,
+      "step": 34310
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 8.891765341970363e-06,
+      "loss": 1.8004,
+      "step": 34320
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 8.876949830644432e-06,
+      "loss": 1.8028,
+      "step": 34330
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 8.862144528951194e-06,
+      "loss": 1.7938,
+      "step": 34340
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 8.84734944404666e-06,
+      "loss": 1.8146,
+      "step": 34350
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 8.832564583081906e-06,
+      "loss": 1.7881,
+      "step": 34360
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 8.81778995320306e-06,
+      "loss": 1.7841,
+      "step": 34370
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 8.803025561551289e-06,
+      "loss": 1.8205,
+      "step": 34380
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 8.78827141526284e-06,
+      "loss": 1.7638,
+      "step": 34390
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 8.773527521468994e-06,
+      "loss": 1.7943,
+      "step": 34400
+    },
+    {
+      "epoch": 1.51,
+      "eval_loss": 1.7974566221237183,
+      "eval_runtime": 11.745,
+      "eval_samples_per_second": 348.745,
+      "eval_steps_per_second": 21.797,
+      "step": 34400
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 8.758793887296068e-06,
+      "loss": 1.7452,
+      "step": 34410
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 8.74407051986544e-06,
+      "loss": 1.8075,
+      "step": 34420
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 8.729357426293494e-06,
+      "loss": 1.8142,
+      "step": 34430
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 8.7146546136917e-06,
+      "loss": 1.8146,
+      "step": 34440
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 8.699962089166505e-06,
+      "loss": 1.8135,
+      "step": 34450
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 8.685279859819418e-06,
+      "loss": 1.7831,
+      "step": 34460
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 8.670607932746948e-06,
+      "loss": 1.8292,
+      "step": 34470
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 8.655946315040642e-06,
+      "loss": 1.7835,
+      "step": 34480
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 8.64129501378706e-06,
+      "loss": 1.8169,
+      "step": 34490
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 8.626654036067783e-06,
+      "loss": 1.8659,
+      "step": 34500
+    },
+    {
+      "epoch": 1.51,
+      "eval_loss": 1.797457218170166,
+      "eval_runtime": 11.691,
+      "eval_samples_per_second": 350.355,
+      "eval_steps_per_second": 21.897,
+      "step": 34500
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 8.612023388959377e-06,
+      "loss": 1.8225,
+      "step": 34510
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 8.597403079533434e-06,
+      "loss": 1.8664,
+      "step": 34520
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 8.582793114856554e-06,
+      "loss": 1.7656,
+      "step": 34530
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 8.568193501990337e-06,
+      "loss": 1.8224,
+      "step": 34540
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 8.55360424799135e-06,
+      "loss": 1.8056,
+      "step": 34550
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 8.539025359911197e-06,
+      "loss": 1.8298,
+      "step": 34560
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 8.52445684479642e-06,
+      "loss": 1.79,
+      "step": 34570
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 8.509898709688614e-06,
+      "loss": 1.7757,
+      "step": 34580
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 8.495350961624296e-06,
+      "loss": 1.8096,
+      "step": 34590
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 8.480813607635003e-06,
+      "loss": 1.7909,
+      "step": 34600
+    },
+    {
+      "epoch": 1.51,
+      "eval_loss": 1.797386884689331,
+      "eval_runtime": 12.0435,
+      "eval_samples_per_second": 340.101,
+      "eval_steps_per_second": 21.256,
+      "step": 34600
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 8.466286654747212e-06,
+      "loss": 1.8285,
+      "step": 34610
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 8.451770109982406e-06,
+      "loss": 1.8151,
+      "step": 34620
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 8.437263980357028e-06,
+      "loss": 1.8071,
+      "step": 34630
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 8.422768272882468e-06,
+      "loss": 1.7474,
+      "step": 34640
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 8.408282994565098e-06,
+      "loss": 1.7808,
+      "step": 34650
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 8.39380815240625e-06,
+      "loss": 1.8051,
+      "step": 34660
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 8.379343753402214e-06,
+      "loss": 1.8293,
+      "step": 34670
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 8.364889804544204e-06,
+      "loss": 1.7951,
+      "step": 34680
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 8.350446312818424e-06,
+      "loss": 1.8187,
+      "step": 34690
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 8.336013285205977e-06,
+      "loss": 1.7762,
+      "step": 34700
+    },
+    {
+      "epoch": 1.52,
+      "eval_loss": 1.7973322868347168,
+      "eval_runtime": 11.7769,
+      "eval_samples_per_second": 347.801,
+      "eval_steps_per_second": 21.738,
+      "step": 34700
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 8.321590728682966e-06,
+      "loss": 1.7932,
+      "step": 34710
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 8.307178650220382e-06,
+      "loss": 1.7925,
+      "step": 34720
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 8.29277705678418e-06,
+      "loss": 1.7959,
+      "step": 34730
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 8.27838595533523e-06,
+      "loss": 1.802,
+      "step": 34740
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 8.26400535282934e-06,
+      "loss": 1.8553,
+      "step": 34750
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 8.249635256217243e-06,
+      "loss": 1.7845,
+      "step": 34760
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 8.23527567244461e-06,
+      "loss": 1.783,
+      "step": 34770
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 8.220926608451988e-06,
+      "loss": 1.8221,
+      "step": 34780
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 8.206588071174878e-06,
+      "loss": 1.7731,
+      "step": 34790
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 8.192260067543685e-06,
+      "loss": 1.8195,
+      "step": 34800
+    },
+    {
+      "epoch": 1.52,
+      "eval_loss": 1.797337293624878,
+      "eval_runtime": 11.9416,
+      "eval_samples_per_second": 343.003,
+      "eval_steps_per_second": 21.438,
+      "step": 34800
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 8.177942604483716e-06,
+      "loss": 1.7583,
+      "step": 34810
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 8.163635688915178e-06,
+      "loss": 1.8383,
+      "step": 34820
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 8.149339327753199e-06,
+      "loss": 1.783,
+      "step": 34830
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 8.13505352790777e-06,
+      "loss": 1.848,
+      "step": 34840
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 8.12077829628384e-06,
+      "loss": 1.8442,
+      "step": 34850
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 8.106513639781188e-06,
+      "loss": 1.8794,
+      "step": 34860
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 8.0922595652945e-06,
+      "loss": 1.7914,
+      "step": 34870
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 8.078016079713358e-06,
+      "loss": 1.7973,
+      "step": 34880
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 8.063783189922223e-06,
+      "loss": 1.7731,
+      "step": 34890
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 8.049560902800438e-06,
+      "loss": 1.8355,
+      "step": 34900
+    },
+    {
+      "epoch": 1.53,
+      "eval_loss": 1.7973167896270752,
+      "eval_runtime": 11.7033,
+      "eval_samples_per_second": 349.986,
+      "eval_steps_per_second": 21.874,
+      "step": 34900
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 8.035349225222201e-06,
+      "loss": 1.8303,
+      "step": 34910
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 8.021148164056604e-06,
+      "loss": 1.7822,
+      "step": 34920
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 8.006957726167596e-06,
+      "loss": 1.8136,
+      "step": 34930
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 7.992777918414007e-06,
+      "loss": 1.8305,
+      "step": 34940
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 7.978608747649503e-06,
+      "loss": 1.7985,
+      "step": 34950
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 7.964450220722632e-06,
+      "loss": 1.8094,
+      "step": 34960
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 7.95030234447677e-06,
+      "loss": 1.845,
+      "step": 34970
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 7.936165125750192e-06,
+      "loss": 1.8332,
+      "step": 34980
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 7.92203857137597e-06,
+      "loss": 1.8231,
+      "step": 34990
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 7.907922688182065e-06,
+      "loss": 1.7771,
+      "step": 35000
+    },
+    {
+      "epoch": 1.53,
+      "eval_loss": 1.7973071336746216,
+      "eval_runtime": 11.7574,
+      "eval_samples_per_second": 348.375,
+      "eval_steps_per_second": 21.773,
+      "step": 35000
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 7.893817482991233e-06,
+      "loss": 1.8482,
+      "step": 35010
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 7.879722962621117e-06,
+      "loss": 1.8413,
+      "step": 35020
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 7.865639133884164e-06,
+      "loss": 1.7733,
+      "step": 35030
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 7.851566003587677e-06,
+      "loss": 1.786,
+      "step": 35040
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 7.837503578533756e-06,
+      "loss": 1.7989,
+      "step": 35050
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 7.823451865519353e-06,
+      "loss": 1.8485,
+      "step": 35060
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 7.809410871336236e-06,
+      "loss": 1.8469,
+      "step": 35070
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 7.795380602771e-06,
+      "loss": 1.7898,
+      "step": 35080
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 7.781361066605038e-06,
+      "loss": 1.774,
+      "step": 35090
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 7.767352269614553e-06,
+      "loss": 1.8445,
+      "step": 35100
+    },
+    {
+      "epoch": 1.54,
+      "eval_loss": 1.7972954511642456,
+      "eval_runtime": 11.9038,
+      "eval_samples_per_second": 344.093,
+      "eval_steps_per_second": 21.506,
+      "step": 35100
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 7.75335421857058e-06,
+      "loss": 1.7657,
+      "step": 35110
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 7.73936692023895e-06,
+      "loss": 1.8643,
+      "step": 35120
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 7.725390381380298e-06,
+      "loss": 1.8439,
+      "step": 35130
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 7.711424608750048e-06,
+      "loss": 1.8153,
+      "step": 35140
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 7.697469609098427e-06,
+      "loss": 1.8191,
+      "step": 35150
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 7.683525389170458e-06,
+      "loss": 1.7889,
+      "step": 35160
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 7.66959195570597e-06,
+      "loss": 1.7851,
+      "step": 35170
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 7.655669315439536e-06,
+      "loss": 1.8457,
+      "step": 35180
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 7.641757475100542e-06,
+      "loss": 1.8443,
+      "step": 35190
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 7.6278564414131515e-06,
+      "loss": 1.7745,
+      "step": 35200
+    },
+    {
+      "epoch": 1.54,
+      "eval_loss": 1.7972785234451294,
+      "eval_runtime": 11.7681,
+      "eval_samples_per_second": 348.059,
+      "eval_steps_per_second": 21.754,
+      "step": 35200
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 7.613966221096313e-06,
+      "loss": 1.7634,
+      "step": 35210
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 7.600086820863717e-06,
+      "loss": 1.8021,
+      "step": 35220
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 7.586218247423864e-06,
+      "loss": 1.7733,
+      "step": 35230
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 7.572360507479981e-06,
+      "loss": 1.841,
+      "step": 35240
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 7.5585136077300895e-06,
+      "loss": 1.8506,
+      "step": 35250
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 7.544677554866964e-06,
+      "loss": 1.7568,
+      "step": 35260
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 7.530852355578142e-06,
+      "loss": 1.8283,
+      "step": 35270
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 7.51703801654589e-06,
+      "loss": 1.8275,
+      "step": 35280
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 7.503234544447244e-06,
+      "loss": 1.8043,
+      "step": 35290
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 7.489441945953994e-06,
+      "loss": 1.8329,
+      "step": 35300
+    },
+    {
+      "epoch": 1.55,
+      "eval_loss": 1.7972854375839233,
+      "eval_runtime": 11.6885,
+      "eval_samples_per_second": 350.431,
+      "eval_steps_per_second": 21.902,
+      "step": 35300
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 7.475660227732672e-06,
+      "loss": 1.8416,
+      "step": 35310
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 7.461889396444537e-06,
+      "loss": 1.8516,
+      "step": 35320
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 7.448129458745573e-06,
+      "loss": 1.8768,
+      "step": 35330
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 7.434380421286559e-06,
+      "loss": 1.8093,
+      "step": 35340
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 7.420642290712938e-06,
+      "loss": 1.8084,
+      "step": 35350
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 7.4069150736649295e-06,
+      "loss": 1.8161,
+      "step": 35360
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 7.393198776777437e-06,
+      "loss": 1.8265,
+      "step": 35370
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 7.379493406680117e-06,
+      "loss": 1.7728,
+      "step": 35380
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 7.365798969997334e-06,
+      "loss": 1.8431,
+      "step": 35390
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 7.35211547334818e-06,
+      "loss": 1.8254,
+      "step": 35400
+    },
+    {
+      "epoch": 1.55,
+      "eval_loss": 1.7973430156707764,
+      "eval_runtime": 11.8006,
+      "eval_samples_per_second": 347.1,
+      "eval_steps_per_second": 21.694,
+      "step": 35400
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 7.338442923346428e-06,
+      "loss": 1.8339,
+      "step": 35410
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 7.324781326600588e-06,
+      "loss": 1.8182,
+      "step": 35420
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 7.311130689713868e-06,
+      "loss": 1.8214,
+      "step": 35430
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 7.297491019284188e-06,
+      "loss": 1.8257,
+      "step": 35440
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 7.28386232190414e-06,
+      "loss": 1.8534,
+      "step": 35450
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 7.2702446041610446e-06,
+      "loss": 1.8448,
+      "step": 35460
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 7.2566378726368776e-06,
+      "loss": 1.783,
+      "step": 35470
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 7.243042133908364e-06,
+      "loss": 1.8584,
+      "step": 35480
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 7.2294573945468515e-06,
+      "loss": 1.826,
+      "step": 35490
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 7.215883661118413e-06,
+      "loss": 1.8125,
+      "step": 35500
+    },
+    {
+      "epoch": 1.55,
+      "eval_loss": 1.7973475456237793,
+      "eval_runtime": 11.7172,
+      "eval_samples_per_second": 349.572,
+      "eval_steps_per_second": 21.848,
+      "step": 35500
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 7.202320940183779e-06,
+      "loss": 1.8274,
+      "step": 35510
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 7.1887692382983675e-06,
+      "loss": 1.7883,
+      "step": 35520
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 7.175228562012272e-06,
+      "loss": 1.8196,
+      "step": 35530
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 7.161698917870266e-06,
+      "loss": 1.802,
+      "step": 35540
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 7.148180312411761e-06,
+      "loss": 1.786,
+      "step": 35550
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 7.134672752170856e-06,
+      "loss": 1.787,
+      "step": 35560
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 7.121176243676315e-06,
+      "loss": 1.802,
+      "step": 35570
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 7.107690793451538e-06,
+      "loss": 1.8255,
+      "step": 35580
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 7.094216408014606e-06,
+      "loss": 1.8201,
+      "step": 35590
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 7.0807530938782216e-06,
+      "loss": 1.79,
+      "step": 35600
+    },
+    {
+      "epoch": 1.56,
+      "eval_loss": 1.7972863912582397,
+      "eval_runtime": 11.965,
+      "eval_samples_per_second": 342.331,
+      "eval_steps_per_second": 21.396,
+      "step": 35600
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 7.067300857549785e-06,
+      "loss": 1.8406,
+      "step": 35610
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 7.05385970553128e-06,
+      "loss": 1.8449,
+      "step": 35620
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 7.04042964431939e-06,
+      "loss": 1.7923,
+      "step": 35630
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 7.027010680405389e-06,
+      "loss": 1.7873,
+      "step": 35640
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 7.0136028202752244e-06,
+      "loss": 1.8115,
+      "step": 35650
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 7.000206070409455e-06,
+      "loss": 1.8063,
+      "step": 35660
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 6.986820437283291e-06,
+      "loss": 1.8808,
+      "step": 35670
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 6.973445927366538e-06,
+      "loss": 1.8388,
+      "step": 35680
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 6.96008254712365e-06,
+      "loss": 1.8139,
+      "step": 35690
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 6.946730303013694e-06,
+      "loss": 1.8185,
+      "step": 35700
+    },
+    {
+      "epoch": 1.56,
+      "eval_loss": 1.7972298860549927,
+      "eval_runtime": 11.8804,
+      "eval_samples_per_second": 344.768,
+      "eval_steps_per_second": 21.548,
+      "step": 35700
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 6.933389201490366e-06,
+      "loss": 1.8188,
+      "step": 35710
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 6.920059249001942e-06,
+      "loss": 1.8225,
+      "step": 35720
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 6.906740451991357e-06,
+      "loss": 1.8263,
+      "step": 35730
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 6.8934328168961e-06,
+      "loss": 1.8524,
+      "step": 35740
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 6.880136350148327e-06,
+      "loss": 1.7724,
+      "step": 35750
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 6.866851058174743e-06,
+      "loss": 1.8175,
+      "step": 35760
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 6.853576947396683e-06,
+      "loss": 1.8516,
+      "step": 35770
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 6.8403140242300554e-06,
+      "loss": 1.7983,
+      "step": 35780
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 6.827062295085378e-06,
+      "loss": 1.8137,
+      "step": 35790
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 6.813821766367759e-06,
+      "loss": 1.7658,
+      "step": 35800
+    },
+    {
+      "epoch": 1.57,
+      "eval_loss": 1.7972460985183716,
+      "eval_runtime": 11.7601,
+      "eval_samples_per_second": 348.295,
+      "eval_steps_per_second": 21.768,
+      "step": 35800
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 6.800592444476874e-06,
+      "loss": 1.8095,
+      "step": 35810
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 6.787374335807001e-06,
+      "loss": 1.7734,
+      "step": 35820
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 6.774167446746992e-06,
+      "loss": 1.8284,
+      "step": 35830
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 6.760971783680281e-06,
+      "loss": 1.8354,
+      "step": 35840
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 6.74778735298486e-06,
+      "loss": 1.803,
+      "step": 35850
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 6.7346141610333145e-06,
+      "loss": 1.8212,
+      "step": 35860
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 6.721452214192765e-06,
+      "loss": 1.7941,
+      "step": 35870
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 6.708301518824949e-06,
+      "loss": 1.767,
+      "step": 35880
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 6.695162081286108e-06,
+      "loss": 1.8077,
+      "step": 35890
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 6.682033907927087e-06,
+      "loss": 1.808,
+      "step": 35900
+    },
+    {
+      "epoch": 1.57,
+      "eval_loss": 1.7970843315124512,
+      "eval_runtime": 11.7297,
+      "eval_samples_per_second": 349.198,
+      "eval_steps_per_second": 21.825,
+      "step": 35900
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 6.6689170050932545e-06,
+      "loss": 1.8053,
+      "step": 35910
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 6.6558113791245475e-06,
+      "loss": 1.8368,
+      "step": 35920
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 6.642717036355456e-06,
+      "loss": 1.817,
+      "step": 35930
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 6.629633983115015e-06,
+      "loss": 1.7996,
+      "step": 35940
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 6.616562225726782e-06,
+      "loss": 1.8287,
+      "step": 35950
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 6.603501770508885e-06,
+      "loss": 1.8236,
+      "step": 35960
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 6.590452623773968e-06,
+      "loss": 1.7861,
+      "step": 35970
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 6.577414791829226e-06,
+      "loss": 1.7758,
+      "step": 35980
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 6.564388280976365e-06,
+      "loss": 1.7861,
+      "step": 35990
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 6.551373097511637e-06,
+      "loss": 1.812,
+      "step": 36000
+    },
+    {
+      "epoch": 1.58,
+      "eval_loss": 1.797182559967041,
+      "eval_runtime": 11.8246,
+      "eval_samples_per_second": 346.396,
+      "eval_steps_per_second": 21.65,
+      "step": 36000
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 6.538369247725795e-06,
+      "loss": 1.8064,
+      "step": 36010
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 6.52537673790416e-06,
+      "loss": 1.8005,
+      "step": 36020
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 6.512395574326524e-06,
+      "loss": 1.8345,
+      "step": 36030
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 6.49942576326721e-06,
+      "loss": 1.8291,
+      "step": 36040
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 6.486467310995062e-06,
+      "loss": 1.8399,
+      "step": 36050
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 6.473520223773433e-06,
+      "loss": 1.8127,
+      "step": 36060
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 6.460584507860181e-06,
+      "loss": 1.8742,
+      "step": 36070
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 6.447660169507653e-06,
+      "loss": 1.8144,
+      "step": 36080
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 6.434747214962718e-06,
+      "loss": 1.8517,
+      "step": 36090
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 6.421845650466734e-06,
+      "loss": 1.8358,
+      "step": 36100
+    },
+    {
+      "epoch": 1.58,
+      "eval_loss": 1.79728364944458,
+      "eval_runtime": 11.953,
+      "eval_samples_per_second": 342.677,
+      "eval_steps_per_second": 21.417,
+      "step": 36100
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 6.408955482255563e-06,
+      "loss": 1.7886,
+      "step": 36110
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 6.39607671655953e-06,
+      "loss": 1.8088,
+      "step": 36120
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 6.3832093596034905e-06,
+      "loss": 1.7968,
+      "step": 36130
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 6.3703534176067415e-06,
+      "loss": 1.8042,
+      "step": 36140
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 6.357508896783093e-06,
+      "loss": 1.8517,
+      "step": 36150
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 6.344675803340829e-06,
+      "loss": 1.8469,
+      "step": 36160
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 6.331854143482715e-06,
+      "loss": 1.7896,
+      "step": 36170
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 6.3190439234059695e-06,
+      "loss": 1.804,
+      "step": 36180
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 6.306245149302297e-06,
+      "loss": 1.8063,
+      "step": 36190
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 6.293457827357871e-06,
+      "loss": 1.8314,
+      "step": 36200
+    },
+    {
+      "epoch": 1.58,
+      "eval_loss": 1.7972148656845093,
+      "eval_runtime": 11.4888,
+      "eval_samples_per_second": 356.521,
+      "eval_steps_per_second": 22.283,
+      "step": 36200
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 6.2806819637533365e-06,
+      "loss": 1.8709,
+      "step": 36210
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 6.267917564663768e-06,
+      "loss": 1.7827,
+      "step": 36220
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 6.255164636258737e-06,
+      "loss": 1.7999,
+      "step": 36230
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 6.242423184702246e-06,
+      "loss": 1.8188,
+      "step": 36240
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 6.229693216152774e-06,
+      "loss": 1.831,
+      "step": 36250
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 6.216974736763212e-06,
+      "loss": 1.8664,
+      "step": 36260
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 6.204267752680943e-06,
+      "loss": 1.7857,
+      "step": 36270
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 6.191572270047754e-06,
+      "loss": 1.8188,
+      "step": 36280
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 6.178888294999892e-06,
+      "loss": 1.8396,
+      "step": 36290
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 6.166215833668055e-06,
+      "loss": 1.8235,
+      "step": 36300
+    },
+    {
+      "epoch": 1.59,
+      "eval_loss": 1.7972338199615479,
+      "eval_runtime": 11.7825,
+      "eval_samples_per_second": 347.634,
+      "eval_steps_per_second": 21.727,
+      "step": 36300
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 6.1535548921773425e-06,
+      "loss": 1.8227,
+      "step": 36310
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 6.140905476647312e-06,
+      "loss": 1.8282,
+      "step": 36320
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 6.128267593191946e-06,
+      "loss": 1.8089,
+      "step": 36330
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 6.115641247919653e-06,
+      "loss": 1.8254,
+      "step": 36340
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 6.103026446933244e-06,
+      "loss": 1.8223,
+      "step": 36350
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 6.090423196329991e-06,
+      "loss": 1.8172,
+      "step": 36360
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 6.077831502201529e-06,
+      "loss": 1.7998,
+      "step": 36370
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 6.0652513706339726e-06,
+      "loss": 1.848,
+      "step": 36380
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 6.052682807707791e-06,
+      "loss": 1.8263,
+      "step": 36390
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 6.040125819497895e-06,
+      "loss": 1.8287,
+      "step": 36400
+    },
+    {
+      "epoch": 1.59,
+      "eval_loss": 1.7972549200057983,
+      "eval_runtime": 11.8176,
+      "eval_samples_per_second": 346.601,
+      "eval_steps_per_second": 21.663,
+      "step": 36400
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 6.027580412073575e-06,
+      "loss": 1.8571,
+      "step": 36410
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 6.015046591498549e-06,
+      "loss": 1.8373,
+      "step": 36420
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 6.002524363830925e-06,
+      "loss": 1.831,
+      "step": 36430
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.990013735123212e-06,
+      "loss": 1.7875,
+      "step": 36440
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.977514711422295e-06,
+      "loss": 1.8114,
+      "step": 36450
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.965027298769468e-06,
+      "loss": 1.8116,
+      "step": 36460
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.952551503200405e-06,
+      "loss": 1.8001,
+      "step": 36470
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.940087330745181e-06,
+      "loss": 1.8408,
+      "step": 36480
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.927634787428219e-06,
+      "loss": 1.82,
+      "step": 36490
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.9151938792683545e-06,
+      "loss": 1.8077,
+      "step": 36500
+    },
+    {
+      "epoch": 1.6,
+      "eval_loss": 1.7971975803375244,
+      "eval_runtime": 12.6073,
+      "eval_samples_per_second": 324.89,
+      "eval_steps_per_second": 20.306,
+      "step": 36500
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.902764612278788e-06,
+      "loss": 1.8095,
+      "step": 36510
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.890346992467084e-06,
+      "loss": 1.8333,
+      "step": 36520
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.877941025835194e-06,
+      "loss": 1.8566,
+      "step": 36530
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.865546718379412e-06,
+      "loss": 1.8239,
+      "step": 36540
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.853164076090429e-06,
+      "loss": 1.8163,
+      "step": 36550
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.840793104953275e-06,
+      "loss": 1.7902,
+      "step": 36560
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.8284338109473526e-06,
+      "loss": 1.8278,
+      "step": 36570
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.8160862000464e-06,
+      "loss": 1.8097,
+      "step": 36580
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.80375027821853e-06,
+      "loss": 1.8486,
+      "step": 36590
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.791426051426197e-06,
+      "loss": 1.8051,
+      "step": 36600
+    },
+    {
+      "epoch": 1.6,
+      "eval_loss": 1.7971279621124268,
+      "eval_runtime": 11.9768,
+      "eval_samples_per_second": 341.994,
+      "eval_steps_per_second": 21.375,
+      "step": 36600
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.779113525626213e-06,
+      "loss": 1.8213,
+      "step": 36610
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.766812706769703e-06,
+      "loss": 1.7921,
+      "step": 36620
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.754523600802175e-06,
+      "loss": 1.8111,
+      "step": 36630
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.742246213663435e-06,
+      "loss": 1.8006,
+      "step": 36640
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.729980551287668e-06,
+      "loss": 1.8423,
+      "step": 36650
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.717726619603353e-06,
+      "loss": 1.7913,
+      "step": 36660
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5.705484424533332e-06,
+      "loss": 1.7874,
+      "step": 36670
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5.693253971994737e-06,
+      "loss": 1.7997,
+      "step": 36680
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5.681035267899055e-06,
+      "loss": 1.8268,
+      "step": 36690
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5.668828318152088e-06,
+      "loss": 1.8094,
+      "step": 36700
+    },
+    {
+      "epoch": 1.61,
+      "eval_loss": 1.797187328338623,
+      "eval_runtime": 11.9607,
+      "eval_samples_per_second": 342.454,
+      "eval_steps_per_second": 21.403,
+      "step": 36700
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5.656633128653958e-06,
+      "loss": 1.8127,
+      "step": 36710
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5.6444497052990805e-06,
+      "loss": 1.8294,
+      "step": 36720
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5.632278053976213e-06,
+      "loss": 1.8017,
+      "step": 36730
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5.6201181805684185e-06,
+      "loss": 1.8082,
+      "step": 36740
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5.6079700909530426e-06,
+      "loss": 1.8064,
+      "step": 36750
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5.595833791001772e-06,
+      "loss": 1.7524,
+      "step": 36760
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5.583709286580551e-06,
+      "loss": 1.7921,
+      "step": 36770
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5.571596583549677e-06,
+      "loss": 1.8524,
+      "step": 36780
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5.559495687763696e-06,
+      "loss": 1.7973,
+      "step": 36790
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5.547406605071474e-06,
+      "loss": 1.8229,
+      "step": 36800
+    },
+    {
+      "epoch": 1.61,
+      "eval_loss": 1.797210931777954,
+      "eval_runtime": 11.794,
+      "eval_samples_per_second": 347.297,
+      "eval_steps_per_second": 21.706,
+      "step": 36800
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5.535329341316149e-06,
+      "loss": 1.8275,
+      "step": 36810
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5.523263902335156e-06,
+      "loss": 1.8013,
+      "step": 36820
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5.511210293960222e-06,
+      "loss": 1.8436,
+      "step": 36830
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5.499168522017351e-06,
+      "loss": 1.7926,
+      "step": 36840
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5.4871385923268105e-06,
+      "loss": 1.8132,
+      "step": 36850
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5.475120510703163e-06,
+      "loss": 1.8226,
+      "step": 36860
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5.463114282955241e-06,
+      "loss": 1.7857,
+      "step": 36870
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5.451119914886143e-06,
+      "loss": 1.8093,
+      "step": 36880
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5.439137412293235e-06,
+      "loss": 1.8043,
+      "step": 36890
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5.427166780968155e-06,
+      "loss": 1.7664,
+      "step": 36900
+    },
+    {
+      "epoch": 1.62,
+      "eval_loss": 1.7971539497375488,
+      "eval_runtime": 11.76,
+      "eval_samples_per_second": 348.299,
+      "eval_steps_per_second": 21.769,
+      "step": 36900
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5.415208026696778e-06,
+      "loss": 1.8128,
+      "step": 36910
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5.403261155259289e-06,
+      "loss": 1.793,
+      "step": 36920
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5.391326172430078e-06,
+      "loss": 1.7863,
+      "step": 36930
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5.379403083977823e-06,
+      "loss": 1.8186,
+      "step": 36940
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5.367491895665422e-06,
+      "loss": 1.8128,
+      "step": 36950
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5.355592613250052e-06,
+      "loss": 1.84,
+      "step": 36960
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5.343705242483113e-06,
+      "loss": 1.8215,
+      "step": 36970
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5.331829789110276e-06,
+      "loss": 1.8136,
+      "step": 36980
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5.319966258871409e-06,
+      "loss": 1.8243,
+      "step": 36990
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5.308114657500649e-06,
+      "loss": 1.8066,
+      "step": 37000
+    },
+    {
+      "epoch": 1.62,
+      "eval_loss": 1.7971651554107666,
+      "eval_runtime": 11.8992,
+      "eval_samples_per_second": 344.224,
+      "eval_steps_per_second": 21.514,
+      "step": 37000
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5.2962749907263694e-06,
+      "loss": 1.8203,
+      "step": 37010
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5.284447264271147e-06,
+      "loss": 1.8284,
+      "step": 37020
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5.272631483851822e-06,
+      "loss": 1.8251,
+      "step": 37030
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5.260827655179419e-06,
+      "loss": 1.8302,
+      "step": 37040
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5.249035783959241e-06,
+      "loss": 1.808,
+      "step": 37050
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5.237255875890762e-06,
+      "loss": 1.8063,
+      "step": 37060
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5.225487936667701e-06,
+      "loss": 1.8649,
+      "step": 37070
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5.2137319719779755e-06,
+      "loss": 1.8114,
+      "step": 37080
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5.2019879875037255e-06,
+      "loss": 1.839,
+      "step": 37090
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5.190255988921298e-06,
+      "loss": 1.8004,
+      "step": 37100
+    },
+    {
+      "epoch": 1.62,
+      "eval_loss": 1.7972087860107422,
+      "eval_runtime": 11.9296,
+      "eval_samples_per_second": 343.347,
+      "eval_steps_per_second": 21.459,
+      "step": 37100
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5.178535981901261e-06,
+      "loss": 1.7963,
+      "step": 37110
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5.166827972108349e-06,
+      "loss": 1.7894,
+      "step": 37120
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5.155131965201535e-06,
+      "loss": 1.8562,
+      "step": 37130
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5.143447966833974e-06,
+      "loss": 1.8311,
+      "step": 37140
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5.1317759826530286e-06,
+      "loss": 1.8282,
+      "step": 37150
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5.12011601830023e-06,
+      "loss": 1.8286,
+      "step": 37160
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5.1084680794113245e-06,
+      "loss": 1.7696,
+      "step": 37170
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5.096832171616225e-06,
+      "loss": 1.8051,
+      "step": 37180
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5.08520830053906e-06,
+      "loss": 1.8004,
+      "step": 37190
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5.073596471798108e-06,
+      "loss": 1.8315,
+      "step": 37200
+    },
+    {
+      "epoch": 1.63,
+      "eval_loss": 1.7971980571746826,
+      "eval_runtime": 11.9874,
+      "eval_samples_per_second": 341.693,
+      "eval_steps_per_second": 21.356,
+      "step": 37200
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5.061996691005847e-06,
+      "loss": 1.8219,
+      "step": 37210
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5.0504089637689166e-06,
+      "loss": 1.7975,
+      "step": 37220
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5.03883329568814e-06,
+      "loss": 1.8664,
+      "step": 37230
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5.027269692358522e-06,
+      "loss": 1.8035,
+      "step": 37240
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5.015718159369207e-06,
+      "loss": 1.8155,
+      "step": 37250
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5.00417870230353e-06,
+      "loss": 1.8054,
+      "step": 37260
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.992651326738983e-06,
+      "loss": 1.8541,
+      "step": 37270
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.9811360382472245e-06,
+      "loss": 1.8034,
+      "step": 37280
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.969632842394051e-06,
+      "loss": 1.7881,
+      "step": 37290
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.958141744739436e-06,
+      "loss": 1.8364,
+      "step": 37300
+    },
+    {
+      "epoch": 1.63,
+      "eval_loss": 1.797140121459961,
+      "eval_runtime": 12.2451,
+      "eval_samples_per_second": 334.501,
+      "eval_steps_per_second": 20.906,
+      "step": 37300
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.946662750837485e-06,
+      "loss": 1.8236,
+      "step": 37310
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.935195866236473e-06,
+      "loss": 1.7661,
+      "step": 37320
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.9237410964788124e-06,
+      "loss": 1.8246,
+      "step": 37330
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.9122984471010675e-06,
+      "loss": 1.8118,
+      "step": 37340
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.900867923633923e-06,
+      "loss": 1.8188,
+      "step": 37350
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.889449531602227e-06,
+      "loss": 1.8013,
+      "step": 37360
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.878043276524956e-06,
+      "loss": 1.78,
+      "step": 37370
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.86664916391522e-06,
+      "loss": 1.834,
+      "step": 37380
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.855267199280254e-06,
+      "loss": 1.8049,
+      "step": 37390
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.843897388121422e-06,
+      "loss": 1.7917,
+      "step": 37400
+    },
+    {
+      "epoch": 1.64,
+      "eval_loss": 1.7971751689910889,
+      "eval_runtime": 12.9807,
+      "eval_samples_per_second": 315.545,
+      "eval_steps_per_second": 19.722,
+      "step": 37400
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.832539735934227e-06,
+      "loss": 1.795,
+      "step": 37410
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.821194248208287e-06,
+      "loss": 1.8131,
+      "step": 37420
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.809860930427332e-06,
+      "loss": 1.7924,
+      "step": 37430
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.798539788069227e-06,
+      "loss": 1.8176,
+      "step": 37440
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.787230826605929e-06,
+      "loss": 1.8091,
+      "step": 37450
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.775934051503523e-06,
+      "loss": 1.8302,
+      "step": 37460
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.764649468222221e-06,
+      "loss": 1.7777,
+      "step": 37470
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.753377082216298e-06,
+      "loss": 1.8047,
+      "step": 37480
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.742116898934166e-06,
+      "loss": 1.8524,
+      "step": 37490
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.73086892381834e-06,
+      "loss": 1.8025,
+      "step": 37500
+    },
+    {
+      "epoch": 1.64,
+      "eval_loss": 1.7971476316452026,
+      "eval_runtime": 12.0152,
+      "eval_samples_per_second": 340.901,
+      "eval_steps_per_second": 21.306,
+      "step": 37500
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 45688,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 500,
+  "total_flos": 9.930378174862983e+18,
+  "train_batch_size": 32,
+  "trial_name": null,
+  "trial_params": null
+}